fleetd #111: probe reads the live memberCredentials policy, no hardcoded name list

scripts/probe-member-credentials.sh carried its own hand-maintained NAMES array
(31 names, recorded 2026-08-16), so a name added later to fleetd.yaml's
memberCredentials.known was never checked and the probe still exited 0 with a
clean-looking table. Same drift shape as #114's tool catalogue.

- New dev.ltms.fleet.member.MemberCredentialPolicyView: the single place that
  turns a MemberCredentials policy into names + counts (never a value). Reused
  by Fleetd.reportMemberCredentialsGap (startup log line) and by the new
  GET /member-credentials REST endpoint (FleetApp), so the two can no longer
  drift apart the way the probe and the policy did.
- FleetApp gains one route + handler + a Supplier<MemberCredentialPolicyView>
  constructor param (legacy constructors default to ::absent, so existing call
  sites are unaffected).
- probe-member-credentials.sh now fetches its name list from
  GET /member-credentials instead of carrying one. No local fallback: an
  unreachable daemon, an empty/absent policy, or a knownCount/known[] length
  mismatch all refuse with a non-zero exit rather than silently checking zero
  names. Prints "policy contains N; this run checked N" so the two numbers are
  visibly equal.

CLAUDE.md

@@ -188,6 +188,18 @@ must obey belongs in the charter, not here.
 - **This repo is the bridge.** The daemon is `fleetd`, its MCP mount is `http://127.0.0.1:8765/mcp`,
   and the code behind the rules above is `mcp/FleetMcp` (tools), `auth/Authz` (the role table),
   `mcp/ConnectionIdentity` (connection→role), and `worker/*Launcher` (`REPLY_CHARTER`).
+- **`fleet_profiles`/`fleet_list` report two separate outage states, and they are not the same
+  thing.** *Quarantined* (CB-578) means the backend told us it is out of capacity — a long,
+  1800s-default cooldown. *Cooling off* (fleetd #201/#227) means a profile's credential threw two
+  distinct backend errors (a non-exhaustion failure such as an HTTP 5xx) within 60 seconds — a
+  short, fixed 60s cooldown, not configurable per profile. Each check runs independently, so a
+  profile can show both at once. In the JSON: a cooling profile carries `credentialId` and
+  `coolingOffForSeconds`; a quarantined profile carries `quarantinedForSeconds`; a profile hit by
+  both carries all three fields, and either state alone already sets that profile's `free` to `0`.
+  A `fleet_spawn` naming a cooling-off profile is refused before it ever reaches the backend
+  adapter, with a message naming the credential and the remaining seconds ("cooling off after
+  repeated backend errors") — distinct wording from a quarantine refusal, so don't conflate the
+  two when reading a spawn failure.
 - **Skills available to delegate:** `implementer` (worktree → commit → push → own PR) and
   `reviewer` (scoped review → one structured finding). Name one in every delegation.
 - **Primary-side skills** (not delegation playbooks — a worker cannot use them):
@@ -195,11 +207,22 @@ must obey belongs in the charter, not here.
   `fleets-status` (report every fleet that shares one LavinMQ instance).
 - **Never commit** `.mcp.json` (the primary's local copy, flagged `--skip-worktree`) or `wiki/`
   (a submodule with its own remote).
-- **Flows and the error model** — rendezvous, `fleet_ask`, detached delivery, turn-done fallback —
-  are diagrammed in `docs/MCP-Contract.md` **§6 only**. The rest of that page is a pre-build design
-  doc whose tool names, parameter names and REST paths never caught up with the code, so do not use
-  it as the tool reference (CB-609). Section 6 is kept out of this file because this file loads into
-  every session's context.
+- **A provisioned worktree neutralizes `.mcp.json`, `opencode.json` and `.autoenv`** — the repo's
+  committed copies would otherwise mount the primary's IDE and forge servers (fleetd #134). The
+  worktree's copy of each is a stub, **not** the repo's real file, so a worker that reads one and
+  reports what it found is reporting on the stub. The daemon logs a per-spawn summary, but the
+  worker cannot see that log. From inside its own worktree a worker — or a lead debugging one —
+  reads the list with `git config --worktree --get-all fleet.neutralizedConfig`, and the
+  consequence with `git config --worktree --get fleet.neutralizedConfigNote`. Never brief a worker
+  to edit one of these files: the edit cannot be committed, and it will not tell you so.
+- **Flows and the error model** — rendezvous, `fleet_ask`, detached delivery, the turn-done
+  fallback and status gating — are diagrammed in `docs/MCP-Contract.md`. That page is now flows
+  only: its pre-build tool catalogue, parameter tables and REST paths were deleted rather than
+  corrected, because a hand-maintained second copy of the tool surface is what drifted for a month
+  while this line pointed every session at it (CB-609 / #114). **The live MCP schema is the tool
+  reference**, with the intent→tool table above as the short form. `McpContractDocTest` fails if
+  that page names a `fleet_*` tool the server does not register. The flows are kept out of this
+  file because this file loads into every session's context.
 
 ### Redeploying the daemon — the lead may do this (primary only)
 

docs/MCP-Contract.md

@@ -1,311 +1,162 @@
-# MCP Contract — `fleetd`'s unified gateway
+# MCP flows and error model — `fleetd`
 
-> **Status: 🔴 HISTORICAL DESIGN — do NOT use as the tool reference.** Written 2026-07-14, before
-> any MCP code existed. The system shipped and this page never caught up, so **its tool names,
-> parameter names and REST paths are wrong today**. Audited 2026-08-17; the specific drift:
+> **What this page is.** The **flows**: how a delegation, a clarification, a detached task and a
+> silent member each travel through `fleetd`. These shapes are what shipped, and they are hard to
+> read off the code because they span the MCP face, the rendezvous registry, the `Injector` and
+> herdr.
 >
-> - **Tools it names that do not exist:** `fleet_read`, `fleet_cancel`.
-> - **Shipped tools it omits:** `fleet_poll`, `fleet_ack`, `fleet_profiles`, `fleet_whoami`.
-> - **Parameter names are wrong nearly everywhere** — it says `message`/`target`/`timeout_seconds`/
->   `block` where the code takes `content`/`sessionId`/`timeoutMs`/`wait`; `text` where
->   `fleet_reply` takes `content`; `target` where `fleet_stop` takes `paneId`.
-> - **REST paths are wrong:** it says `POST /workers` and `DELETE /workers/{paneId}`; the daemon
->   serves `POST /members` and `DELETE /members/{paneId}`.
+> **What this page is NOT: a tool reference.** It deliberately holds no tool catalogue, no
+> parameter tables and no REST paths. **The live MCP schema is the authority** — each tool's own
+> description and parameters, as mounted — with the intent→tool table in `CLAUDE.md` as the short
+> form.
 >
-> **The authoritative tool surface is the live MCP schema** (each tool's own description and
-> parameters, as mounted), with the intent→tool table in `CLAUDE.md` as the short form. Both were
-> checked against `mcp/FleetMcp.java` on 2026-08-17 and are accurate.
+> That absence is the fix for fleetd #114 (CB-609), and it is worth stating why. This page used to
+> carry a full tool catalogue written in July 2026, before any MCP code existed. The code shipped;
+> the page did not follow. By August it named two tools that do not exist, omitted five that do,
+> had the wrong name for nearly every parameter, pointed at REST paths the daemon does not serve,
+> and — worst — still described an identity model (*"any connection that does not map to a known
+> worker is treated as a primary"*) that was a real privilege bug, fixed since by the ancestry
+> walk in fleetd #161. Every one of those errors is the same error: **a second, hand-maintained
+> copy of something the code already states**. So the second copy is gone rather than corrected.
+> Only the flows remain, because a flow is a shape rather than a name, and shapes are what this
+> page was ever good for.
 >
-> What is still worth reading here is **§6 — the flows and the error model** (rendezvous,
-> `fleet_ask`, detached delivery, the turn-done fallback). The shapes it describes are the ones
-> that shipped; only the names around them drifted. Rewriting this page is tracked as **CB-609**.
-
-`fleetd` is the **sole communication gateway** for every Claude session in the bridge. Both
-the **primary** (Opus, on subscription) and every **worker** (off-subscription Claude Code)
-mount the *same* MCP server with a single `claude mcp add` line, and talk only through its
-tools. No Claude session ever addresses a broker, a peer, or the network directly.
-
-This document defines every MCP tool that face must expose, who may call it, its blocking
-semantics, and how it maps onto the code already in the tree.
+> The names that do appear below are checked by `McpContractDocTest`, which fails if this page
+> names a `fleet_*` tool the server does not register. That test is the whole reason it is safe to
+> write a tool name here at all.
 
 ---
 
-## 1. Design constraints (non-negotiable)
+## 1. Rendezvous flows
 
-These come from the project's core invariants and bound every decision below.
+### 1.1 Delegation — happy path
 
-1. **One server, both roles.** The primary and all workers mount an identical server. The
-   catalog must serve both, and `fleetd` must decide *who is calling* from the connection —
-   never from a caller-supplied argument that could be spoofed.
-2. **Subscription-safe by construction.** No MCP tool ever reads, sets, or forwards
-   `ANTHROPIC_BASE_URL`. Mounting the bridge cannot move a session off subscription.
-   Enforced today by [`SubscriptionGuard`](1-Architecture).
-3. **Blocking rendezvous, no busy-poll.** The primary consumes a worker's reply through a
-   *single* MCP call that `fleetd` holds open — never a cross-turn poll loop that would burn
-   subscription quota.
-4. **Status-gated delivery.** Anything that puts text into a worker flows through the existing
-   [`Injector`](1-Architecture): delivered only when the worker is `idle`/`blocked`, at most
-   one message per turn.
-5. **`fleetd` owns policy; herdr owns PTYs.** MCP tools express *intent*; `fleetd`
-   translates it into guard checks, rendezvous bookkeeping, and herdr `agent.*` calls.
-
----
-
-## 2. Topology
-
-Both faces live in the one daemon. The **north face** is MCP (this document); the **south
-face** is the herdr Unix socket. REST/SSE remains only for non-Claude clients and dashboards.
-
-```mermaid
-flowchart LR
-    OPUS["Opus — primary<br/>(Claude Code, env CLEAN)<br/>MCP client"]
-    subgraph BD["fleetd — standalone daemon"]
-        MCP["MCP server (north face)<br/>fleet_send · fleet_reply<br/>fleet_ask · fleet_status · lifecycle"]
-        RDV["rendezvous registry<br/>(blocking-call waiters)"]
-        INJ["Injector + StatusPoller<br/>(status-gated writer)"]
-        SOCK["herdr socket client (south face)"]
-        MCP --> RDV
-        RDV --> INJ
-        INJ --> SOCK
-        MCP --> SOCK
-    end
-    HERDR["herdr<br/>panes · agent-status"]
-    W["worker claude pane<br/>ANTHROPIC_BASE_URL set<br/>MCP client"]
-
-    OPUS -->|"fleet_send (blocks)"| MCP
-    W -.->|"fleet_reply / fleet_ask"| MCP
-    SOCK -->|"agent.start · agent.send<br/>agent.get · pane.close"| HERDR
-    HERDR -->|"drives PTY"| W
-
-    classDef ext fill:#2b6cb0,stroke:#1a365d,color:#ffffff;
-    classDef core fill:#2f855a,stroke:#22543d,color:#ffffff;
-    class OPUS,W ext
-    class MCP,RDV,INJ,SOCK core
-```
-
----
-
-## 3. Identity & addressing
-
-Because the same server is mounted by everyone, `fleetd` resolves the caller's role on every
-request — this is the linchpin of the whole contract and has no code yet.
-
-- **Workers are known.** `fleetd` spawns every worker
-  ([`WorkerService`](1-Architecture)) and records its herdr session UUID / `terminal_id` on
-  the returned [`Agent`]. When a call arrives on a connection that maps to a known worker,
-  the caller is *that* worker — so **workers never pass a target**; routing is implicit.
-- **The primary is "not a worker".** Any connection that does not map to a known worker is
-  treated as a primary. It addresses workers **explicitly** by `target` — a session UUID,
-  a `terminal_id`, or a friendly `profile` name.
-- **Turn correlation.** A blocking `fleet_send` registers a *waiter* keyed by worker
-  identity. A worker's later `fleet_reply` / `fleet_ask` on the same identity resolves that
-  waiter. A `turn_id` is minted per exchange so a clarification round-trip
-  (§6.2) rejoins the right turn.
-
----
-
-## 4. Transport
-
-`fleetd` is a long-lived daemon serving **multiple** concurrent clients (one primary + N
-workers), so a per-client stdio child is the wrong shape. The recommended transport is
-**streamable-HTTP / SSE** on the same bind as the REST face:
-
-```bash
-# identical on primary and every worker
-claude mcp add --transport http fleetd http://127.0.0.1:8080/mcp
-```
-
-This adds an MCP-server dependency the pom does not yet carry. See [Open decisions](#10-open-decisions).
-
----
-
-## 5. Tool catalog
-
-| Tool | Caller | Blocks? | Backing (exists today?) |
-|---|---|---|---|
-| [`fleet_send`](#fleet_send) | primary | yes (default) | `Injector.enqueue` ✅ · rendezvous registry ❌ (CB-104) |
-| [`fleet_reply`](#fleet_reply) | worker | no | rendezvous ❌ · pane injection via `Injector` ✅ |
-| [`fleet_ask`](#fleet_ask) | worker | yes | reverse rendezvous ❌ |
-| [`fleet_status`](#fleet_status) | either | no | `AgentControl.status` ✅ · `Injector.activeTargets` ✅ |
-| [`fleet_spawn`](#lifecycle) | primary | no | `WorkerService.spawn` ✅ (`POST /workers`) |
-| [`fleet_list`](#lifecycle) | either | no | `WorkerService.list` ✅ (`/agents`) |
-| [`fleet_stop`](#lifecycle) | primary | no | `WorkerService.stop` ✅ (`DELETE /workers/{paneId}`) |
-| [`fleet_read`](#fleet_read) | primary | no | `AgentControl.read` ✅ |
-| [`fleet_cancel`](#fleet_cancel) | primary | no | — ❌ (future) |
-
-### Core: delegation & rendezvous
-
-#### `fleet_send`
-*(primary → worker — the headline tool, CB-104)*
-
-- **Params:** `message` (required); `target` (optional — defaults to the sole worker / default
-  profile); `timeout_seconds` (default 600); `block` (default `true`); `auto_spawn`
-  (default `true`); `turn_id` (optional — supplied when answering a worker's `fleet_ask`).
-- **Blocking (`block:true`):** enqueue `message` via the `Injector`, then hold the call open
-  until exactly one of:
-  - worker calls `fleet_reply` → `{ outcome:"reply", text }`
-  - worker calls `fleet_ask` → `{ outcome:"question", text, turn_id }`
-  - worker's `agent_status` reaches done/idle with no reply → `{ outcome:"turn_done", text:<terminal tail> }`
-  - deadline elapses → `{ outcome:"timeout" }`
-  - worker gone → error `worker_gone`
-- **Detached (`block:false`):** enqueue and return `{ outcome:"dispatched", dispatch_id }`
-  immediately. The eventual reply is injected into the primary's idle pane (§6.3), or drained
-  via `fleet_status` on a split-host primary.
-
-#### `fleet_reply`
-*(worker → primary)*
-
-- **Params:** `text` (required); `final` (default `true`).
-- **Behavior:** resolve the primary waiter registered against this worker with `text`. If no
-  waiter exists (detached delegation), `fleetd` **injects the primary's idle pane** instead.
-  Returns `{ delivered:true, mode:"resolved"|"injected" }`. No `target` — identity is implicit.
-
-#### `fleet_ask`
-*(worker → primary — the reverse rendezvous)*
-
-- **Params:** `question` (required); `timeout_seconds`.
-- **Behavior:** blocks the *worker's* call. Surfaces the question to the primary (resolving its
-  open `fleet_send` with `outcome:"question"`, or injecting its pane). When the primary
-  answers — a `fleet_send` carrying the matching `turn_id` — that unblocks this call and
-  returns `{ answer }` to the worker, which continues **in the same turn**.
-
-### Worker lifecycle
-<a id="lifecycle"></a>
-Thin adapters over [`WorkerService`](1-Architecture) — parity with the existing REST routes.
-
-- **`fleet_spawn`** — `{ profile? }` → worker view (`sessionId`, `terminalId`, `paneId`,
-  `status`). Guard-checked; a boundary breach returns error `subscription_boundary` (the
-  REST `403`).
-- **`fleet_list`** — no params → all workers + `agent_status`. Read-only, either role.
-- **`fleet_stop`** — `{ target }` → tears down the pane and its dedicated tab. Idempotent.
-
-### Observability
-
-#### `fleet_status`
-*(either role — the README's 4th named tool)*
-
-- **Params:** `target?`.
-- **Behavior:** per-worker `agent_status`, queue depth (`Injector.activeTargets`), whether a
-  rendezvous is open, and ids. For the *calling* session it also reports/drains **pending
-  messages addressed to me** — the path a split-host primary's `Stop`-hook uses to wake and
-  collect replies without being injectable. Read-only, non-blocking.
-
-#### `fleet_read`
-*(primary)*
-
-- **Params:** `target`; `source` ∈ `visible | recent | recent_unwrapped | detection`.
-- **Behavior:** returns the worker's terminal text so the primary can peek at a *detached*
-  worker's progress. Adapter over `AgentControl.read`.
-
-### Control (future)
-
-#### `fleet_cancel`
-*(primary)*
-
-- **Params:** `target`. Interrupt the worker's current turn / abandon the rendezvous. No
-  backing code yet.
-
----
-
-## 6. Rendezvous flows
-
-### 6.1 Delegation — happy path
-
-One blocking call, zero polls.
+One blocking call, zero polls. The lead's call is held open by `fleetd` until the member answers.
 
 ```mermaid
 sequenceDiagram
-    participant P as Primary (Opus)
-    participant B as fleetd (MCP + Injector)
+    participant P as "Lead (primary)"
+    participant B as "fleetd (MCP + Injector)"
     participant H as herdr
-    participant W as Worker (Claude)
+    participant W as "Member"
 
-    P->>B: fleet_send("do X", target=w) — blocks
-    B->>B: register waiter(w)
-    B->>H: agent.send(w, "do X")  (idle window)
-    H-->>W: prompt injected
-    W->>W: works the turn
-    W->>B: fleet_reply("result")
-    B->>B: resolve waiter(w)
-    B-->>P: { outcome:"reply", text:"result" }
+    P->>B: "fleet_send{sessionId, content} — blocks"
+    B->>B: "register waiter(sessionId)"
+    B->>H: "agent.send — only in an injectable window"
+    H-->>W: "prompt injected"
+    W->>W: "works the turn"
+    W->>B: "fleet_reply{content}"
+    B->>B: "resolve waiter"
+    B-->>P: "{ outcome: reply }"
 ```
 
-### 6.2 Clarification — reverse rendezvous (`fleet_ask`)
+**The cap that matters:** a blocking `fleet_send` is bounded by the *caller's own* MCP client
+timeout, about 60 seconds — not by the task. Anything slower than that must use the detached flow
+in §1.3, or the lead's call returns while the member is still working.
 
-The worker pauses mid-turn to ask; the primary answers; the worker resumes in the same turn.
+### 1.2 Clarification — reverse rendezvous
+
+The member pauses mid-turn to ask, the lead answers, and the member resumes **the same turn** with
+its context intact.
 
 ```mermaid
 sequenceDiagram
-    participant P as Primary
+    participant P as "Lead"
     participant B as fleetd
-    participant W as Worker
+    participant W as "Member"
 
-    P->>B: fleet_send("do X", target=w) — blocks
-    B-->>W: "do X" (injected)
-    W->>B: fleet_ask("which config?") — worker blocks
-    B-->>P: { outcome:"question", text:"which config?", turn_id }
-    P->>B: fleet_send("config.yaml", target=w, turn_id) — blocks again
-    B-->>W: resolve fleet_ask → { answer:"config.yaml" }
-    W->>W: resumes same turn
-    W->>B: fleet_reply("done")
-    B-->>P: { outcome:"reply", text:"done" }
+    P->>B: "fleet_send{sessionId, content} — blocks"
+    B-->>W: "content injected"
+    W->>B: "fleet_ask{question} — member blocks"
+    B-->>P: "{ outcome: question, turnId }"
+    P->>B: "fleet_send{turnId, content} — answers THIS turn"
+    B-->>W: "fleet_ask returns the answer"
+    W->>W: "resumes the same turn"
+    W->>B: "fleet_reply{content}"
+    B-->>P: "{ outcome: reply }"
 ```
 
-### 6.3 Detached delegation — pane injection
+**Answer with `turnId`, never `sessionId`.** A `sessionId` send starts a new turn; it does not
+resolve the waiting `fleet_ask`.
 
-The primary does not block; the reply arrives later in its idle pane.
+**The window is about 55 seconds and no nudge extends it.** So never brief a member to "ask me":
+decide before delegating, or give the member an explicit default to fall back on.
+
+### 1.3 Detached delegation — the lead does not block
+
+The lead gets a ticket immediately and collects the answer later. This is the flow for any real
+task, because of the ~60s cap in §1.1.
 
 ```mermaid
 sequenceDiagram
-    participant P as Primary
+    participant P as "Lead"
     participant B as fleetd
-    participant W as Worker
+    participant W as "Member"
 
-    P->>B: fleet_send("do X", target=w, block=false)
-    B-->>P: { outcome:"dispatched", dispatch_id }
-    P->>P: continues its own work
-    W->>B: fleet_reply("result")
-    Note over B: no waiter → detached path
-    B->>B: Injector.enqueue(primary_pane, "result")
-    B-->>P: injected into idle pane (status-gated)
+    P->>B: "fleet_send{sessionId, content, wait:false}"
+    B-->>P: "accepted — ticket"
+    P->>P: "continues its own work"
+    W->>B: "fleet_reply{content}"
+    Note over B: "no waiter is blocked — the reply is held"
+    B->>B: "nudge the lead's own pane (status-gated)"
+    P->>B: "fleet_poll{ticket}"
+    B-->>P: "the member's report"
+    P->>B: "fleet_ack{target, msgId}"
 ```
 
-### 6.4 Uncooperative worker — turn-done fallback
+A terminal ticket nudges the lead's pane by itself, so a detached task does not need watching. The
+nudge needs an injectable lead pane and is capped, so it is a convenience rather than a guarantee.
 
-A worker that never calls `fleet_reply` still returns a result: `fleetd` reads its terminal
-tail when the turn completes.
+### 1.4 The member never replies — turn-done fallback
+
+A member that ends its turn without `fleet_reply` still produces something: `fleetd` reads its
+pane tail. This is a **fallback, not a channel** — it is lossy in three separate ways, and every
+one of them has produced a wrong answer in practice.
 
 ```mermaid
 sequenceDiagram
-    participant P as Primary
+    participant P as "Lead"
     participant B as fleetd
-    participant W as Worker
+    participant W as "Member"
 
-    P->>B: fleet_send("do X", target=w) — blocks
-    B-->>W: "do X" (injected)
-    W->>W: works, never calls fleet_reply
-    B->>B: StatusPoller sees agent_status → idle/done
-    B->>B: AgentControl.read(w, "recent")
-    B-->>P: { outcome:"turn_done", text:<terminal tail> }
+    P->>B: "fleet_send — blocks or detaches"
+    B-->>W: "content injected"
+    W->>W: "works, never calls fleet_reply"
+    B->>B: "StatusPoller sees the turn end"
+    B->>B: "read the pane tail"
+    B->>B: "classify: exhausted? echoed brief? real report?"
+    B-->>P: "{ outcome: turn_done } or a named failure"
 ```
 
+The three ways it goes wrong, and what each looks like now:
+
+| What happened | What the lead used to get | What it gets today |
+|---|---|---|
+| The report is longer than the scrape window | The **end** silently cut off | Still clipped, but marked partial |
+| The member never started — spent credential | The lead's **own brief** echoed back as a report | A named failure: backend exhausted |
+| The member is simply slow | A tail of work in progress | Unchanged — read it as a hint, not a result |
+
+The echoed-brief case is the one to remember: it reads as a long, on-topic report with nothing in
+it from the member. It is suppressed now, but the general rule stands — **check the member's
+worktree with `git log` before believing a report you did not watch arrive.**
+
 ---
 
-## 7. Status gating
+## 2. Status gating
 
-Delivery only happens in a safe window. This is the state machine the `Injector` already
-enforces via `AgentStatus.injectable()`; MCP `fleet_send` is simply its producer.
+Delivery only happens in a safe window. `fleet_send` is a producer for the `Injector`, which
+already enforces this through `AgentStatus.injectable()`.
 
 ```mermaid
 stateDiagram-v2
     [*] --> IDLE
-    IDLE --> WORKING: message delivered / picks up
-    WORKING --> IDLE: turn done
-    WORKING --> BLOCKED: awaits input
-    BLOCKED --> WORKING: input delivered
-    IDLE --> UNKNOWN: detection glitch
-    BLOCKED --> UNKNOWN: detection glitch
-    UNKNOWN --> IDLE: re-detected
+    IDLE --> WORKING: "message delivered, picked up"
+    WORKING --> IDLE: "turn done"
+    WORKING --> BLOCKED: "awaits input"
+    BLOCKED --> WORKING: "input delivered"
+    IDLE --> UNKNOWN: "detection glitch"
+    BLOCKED --> UNKNOWN: "detection glitch"
+    UNKNOWN --> IDLE: "re-detected"
 
     note right of IDLE
         injectable — deliver head of FIFO
@@ -321,69 +172,17 @@ stateDiagram-v2
     end note
 ```
 
-At most one message is delivered per turn: after a send the `Injector` waits for a `WORKING`
-pickup before delivering the next, with a `PICKUP_GRACE_POLLS` fallback for turns faster than
-the poll interval. A herdr `events.subscribe` stream can later replace the sampling without
-touching this state machine.
+**At most one message per turn.** After a send, the `Injector` waits for a `WORKING` pickup before
+delivering the next, with a grace-poll fallback for turns that finish faster than the poll
+interval.
 
----
+Two consequences a lead feels directly:
 
-## 8. Error model
+- **A second send to a busy member never lands.** It reports as queued and times out. The member
+  is fine; the message simply waits, and then restarts the member when it next goes idle.
+- **A spawned member is not deliverable until it has mounted the MCP.** Until then a send waits on
+  that gate for about 60 seconds and then fails without ever reaching the pane.
 
-| Condition | `fleet_send` result | Notes |
-|---|---|---|
-| Worker replies | `{ outcome:"reply" }` | normal |
-| Worker asks | `{ outcome:"question", turn_id }` | answer with `fleet_send(turn_id)` |
-| Turn ends, no reply | `{ outcome:"turn_done" }` | terminal tail as text |
-| Deadline elapsed | `{ outcome:"timeout" }` | message may still be queued/delivered |
-| Worker vanished | error `worker_gone` | `Injector.drop` fails the queued future |
-| Guard breach on spawn | error `subscription_boundary` | REST `403` parity |
-| Delivery failed at herdr | error, message dropped | poisoned message not left blocking the FIFO |
-
-`fleet_reply` from a worker with no open waiter is **not** an error — it falls through to
-detached pane injection (§6.3).
-
----
-
-## 9. Mapping to existing code
-
-The MCP face is a thin adapter layer; nearly every capability already exists behind the REST
-seam. Only the **rendezvous registry** and the **caller-identity resolver** are new.
-
-| MCP tool | Existing collaborator | New work |
-|---|---|---|
-| `fleet_send` | `Injector.enqueue`, `AgentControl.send` | waiter registry, timeout, outcome mux (CB-104) |
-| `fleet_reply` / `fleet_ask` | `Injector` (pane injection) | reverse rendezvous, identity resolver |
-| `fleet_status` | `AgentControl.status`, `Injector.activeTargets` | pending-drain projection |
-| `fleet_spawn` / `list` / `stop` | `WorkerService.{spawn,list,stop}` | MCP adapter only |
-| `fleet_read` | `AgentControl.read` | MCP adapter only |
-
-Because the REST routes in `FleetApp` already exercise the collaborators, MCP tools are
-validated by **parity** against those routes, not by re-testing behavior.
-
----
-
-## 10. Open decisions
-
-1. **`fleet_ask` direction.** This page defines it as *worker-asks-primary* (a genuine reverse
-   channel, matching the "inject the primary's pane" language). The alternative — a synonym for
-   a blocking primary→worker send — is weaker and produces different plumbing. **Recommend
-   worker-asks-primary.**
-2. **Detached delivery shape.** A `block:false` param on `fleet_send` (keeps the catalog
-   small) vs. a separate `fleet_dispatch` tool. **Recommend the param.**
-3. **Auto-spawn on send.** `fleet_send` provisions a worker per profile when none exists
-   (simplest primary UX) vs. requiring an explicit `fleet_spawn` first. **Recommend
-   auto-spawn, defaulting on.**
-4. **Transport & SDK.** Streamable-HTTP/SSE co-located with the REST bind (recommended) vs.
-   stdio. Requires choosing a Java MCP server SDK and adding it to the pom.
-
----
-
-## 11. Implementation staging
-
-- **CB-104** — blocking `fleet_send` + rendezvous registry + caller-identity resolver
-  (the producer that finally drives the inert `StatusPoller`).
-- **CB-1xx** — `fleet_reply` / `fleet_ask` reverse rendezvous + detached pane injection.
-- **CB-1xx** — lifecycle + observability adapters (`fleet_spawn/list/stop/status/read`).
-- **CB-1xx** — transport wiring + `claude mcp add` docs; parity tests vs. REST.
-- **Later** — `fleet_cancel`; swap `StatusPoller` for herdr `events.subscribe`.
+`UNKNOWN` is deliberately neither injectable nor a pickup. A pane whose status cannot be read is
+not a pane that is safe to write to — see fleetd #176 for what happens when a gate treats an
+unreadable pane as a ready one.

fleetd/fleetd.example.yaml

@@ -172,7 +172,11 @@ herdrSocket: ~/.config/herdr/herdr.sock
 #                      skills/MCP/hooks. Omit to leave the worker on the host default.
 #   parityOverlay    → repo-relative paths copied primary→worktree so a worker in a provisioned
 #                      worktree sees the same local config (CB-301-ext). Omit for the default set:
-#                      [.env, .envrc]. (.claude/settings.local.json is NOT in the default — it
+#                      [.env] only (CB-148). .envrc is left out of the default on purpose: it is
+#                      executable shell that direnv runs on every cd, so copying it carries
+#                      behaviour into the worker, not just values, unlike .env. An operator who
+#                      wants it copied can still write parityOverlay: [.env, .envrc] explicitly.
+#                      (.claude/settings.local.json is NOT in the default — it
 #                      pre-approves IDE/tool grants a member must not hold ambiently; CB-525/CB-634.)
 #
 #                      Do NOT add .mcp.json (CB-525). A worker's tools are whatever its launcher
@@ -206,6 +210,39 @@ herdrSocket: ~/.config/herdr/herdr.sock
 #                      profile quarantines alone, under its own name, exactly as if the field did
 #                      not exist. Cooldown length is the top-level quarantineCooldownSeconds below.
 #                      HOT: read live at every spawn/exhaustion check — no restart needed.
+#   errorPattern     → fleetd #201 / #227: regex matched against a completion-fallback scrape to
+#                      classify a turn that ended with no fleet_reply as a BACKEND ERROR — a
+#                      credential outage or a provider 5xx — rather than a real answer or a
+#                      usage-limit exhaustion (exhaustedPattern above always wins when a line
+#                      matches both). Opt-in. Omit it and this profile falls back to fleetd's
+#                      built-in legacy pattern `(?i)\bAPI Error\s*:` — classification still
+#                      happens, just without a profile-specific match; every backend words its
+#                      failure differently, so a hardcoded sentence would only ever match one
+#                      of them.
+#                      DEFERRED: compiled once into a startup pattern map, same as exhaustedPattern
+#                      — editing it needs a daemon restart.
+#                      # errorPattern: "503 Service Unavailable"   # opt-in: classify a backend outage
+#
+#                      What happens once a match fires (BackendOutagePolicy, credentialId-keyed,
+#                      SEPARATE from the CB-578 stage B quarantine above and never merged with it):
+#                        - threshold 2 — TWO DISTINCT TARGETS (never raw events) on the same
+#                          effective credential inside a 60-second window start an "incident" and a
+#                          60-second cool-off for that credential. One member repeating the same
+#                          classified line twice never cools anything off — a real outage hits
+#                          every target on that credential, so requiring a second, independent
+#                          target loses nothing against the case this guards against, while
+#                          protecting against a heuristic misfire on one flaky member.
+#                        - a fresh error while a credential is already cooling off is ignored
+#                          outright: it neither extends the 60s deadline nor starts a new incident.
+#                        - `fleet_list`/`fleet_profiles` report a cooling credential with
+#                          `coolingOffForSeconds` (never `quarantinedForSeconds`, unless CB-578
+#                          exhaustion quarantine is ALSO independently active for the same
+#                          credential — the two checks can both fire at once). A spawn onto a
+#                          cooling profile is refused with a message naming the credential and
+#                          remaining seconds — "cooling off", never "exhausted", so an operator can
+#                          tell a short transient fault from a spent subscription at a glance.
+#                        - the lead gets ONE nudge per incident (not one per affected target), via
+#                          the same push loop that already delivers ticket/question reminders.
 #   env              → extra environment for this profile's workers, as a literal key/value map
 #                      (CB-511). Use it to give workers a toolchain.
 #
@@ -273,9 +310,10 @@ profiles:
     # gitHostEnv: GITEA_HOST          # defaults to GITEA_HOST; injected only with gitTokenEnv
     # exhaustedPattern: "usage limit has been reached"   # opt-in: classify a usage-limit refusal (CB-578)
     # credentialId: shared-openai   # opt-in: quarantine together with every other profile sharing this id (CB-578)
+    # errorPattern: "503 Service Unavailable"   # opt-in: classify a backend outage (fleetd #201/#227) — see the key doc above
     # configDir: /Users/me/.ccs/instances/gx10   # CLAUDE_CONFIG_DIR — inherit that profile's skills/MCP
     # cwd: /Users/me/src/myrepo       # pin the working dir; omit to inherit the primary's
-    # parityOverlay: [".env", ".envrc"]   # the default; never add .mcp.json or .claude/settings.local.json — see above
+    # parityOverlay: [".env"]   # the default; add ".envrc" explicitly if you want it copied too (CB-148) — never add .mcp.json or .claude/settings.local.json — see above
     # ideMcpUrl: http://127.0.0.1:29170/index-mcp/streamable-http   # opt-in (CB-634): IDE code intelligence, pinned to the worktree
     # ideProjectDir: fleetd          # CB-634: module dir the IDE opens + the overlay pins (this repo's pom is in fleetd/)
     # ideOpenCommand: env DISPLAY=:10.0 idea {dir}   # CB-634 auto-open: opens {dir} in the IDE at spawn; omit to open by hand
@@ -369,6 +407,12 @@ placement: weighted
 # DEFERRED: baked once into the BackendQuarantine built at startup — a running quarantine keeps
 # its original cooldown regardless; a new value only applies to a quarantine that starts after a
 # restart. Editing this needs a daemon restart to take effect.
+#
+# This does NOT govern the fleetd #201 / #227 backend-error cool-off documented under errorPattern
+# above — that mechanism is a separate, shorter-lived, NOT-configurable policy (threshold 2 distinct
+# targets, 60-second window, 60-second cool-off), on purpose: it exists to survive a brief transient
+# fault, not to replace this 30-minute exhaustion quarantine. Do not conflate the two when reading
+# fleet_list/fleet_profiles — coolingOffForSeconds and quarantinedForSeconds are independent facts.
 # quarantineCooldownSeconds: 1800
 
 # Re-read this file without restarting the daemon (CB-559). Off unless you add this block, so an
@@ -395,8 +439,10 @@ placement: weighted
 #               stage B — baked once into the quarantine tracker built at startup), ADDING or
 #               REMOVING a profile (a new backend needs its own launcher, and launchers are built
 #               once), AND an existing profile's launch settings — model, baseUrl, argv, env,
-#               configDir, mcpUrl, tabLabel, exhaustedPattern. The launcher takes a copy of
-#               `profiles:` at startup and resolves every spawn out of that copy, so those never
+#               configDir, mcpUrl, tabLabel, exhaustedPattern, errorPattern (fleetd #201 / #227 —
+#               compiled once into a startup pattern map the same way exhaustedPattern is). The
+#               launcher takes a copy of `profiles:` at startup and resolves every spawn out of
+#               that copy, so those never
 #               reach a launch until you restart. The reload logs them by name rather than
 #               pretending they applied.
 #   COLD      → cannot change at all: `bind:`, `herdrSocket:`, `broker:` and `auth:`. The socket is

fleetd/src/main/java/dev/ltms/fleet/Fleetd.java

@@ -13,6 +13,8 @@ import dev.ltms.fleet.lead.LeadLauncher;
 import dev.ltms.fleet.herdr.PaneLocator;
 import dev.ltms.fleet.herdr.UnixSocketHerdrClient;
 import dev.ltms.fleet.herdr.WorkspaceControl;
+import dev.ltms.fleet.inject.BackendErrorPatternLookup;
+import dev.ltms.fleet.inject.BackendErrorSink;
 import dev.ltms.fleet.inject.CompletionResolver;
 import dev.ltms.fleet.inject.ExhaustedPatternLookup;
 import dev.ltms.fleet.inject.ExhaustionSink;
@@ -49,7 +51,9 @@ import dev.ltms.fleet.session.SessionReaper;
 import dev.ltms.fleet.member.ClaudeCodeLauncher;
 import dev.ltms.fleet.member.CompositePeerLauncher;
 import dev.ltms.fleet.member.HerdrPeerLauncher;
+import dev.ltms.fleet.member.MemberCredentialPolicyView;
 import dev.ltms.fleet.member.OpenCodeLauncher;
+import dev.ltms.fleet.placement.BackendOutagePolicy;
 import dev.ltms.fleet.placement.BackendQuarantine;
 import io.javalin.Javalin;
 import org.slf4j.Logger;
@@ -61,6 +65,7 @@ import java.util.ArrayList;
 import java.util.LinkedHashMap;
 import java.util.List;
 import java.util.Map;
+import java.util.Optional;
 import java.util.Set;
 import java.util.concurrent.Executors;
 import java.util.concurrent.ScheduledExecutorService;
@@ -205,12 +210,18 @@ public final class Fleetd {
         // here, at startup, and a config reload only changes it for a daemon restart.
         BackendQuarantine quarantine = new BackendQuarantine(System::nanoTime,
                 TimeUnit.SECONDS.toNanos(cfg.quarantineCooldownSeconds()));
+        // fleetd #201 Unit 5: one outage-cool-off tracker for the whole daemon, shared between the
+        // launcher (checked at spawn, like `quarantine` above) and the backend-error sink wired in
+        // below (written on a classified backend error). A SEPARATE, shorter-lived mechanism from
+        // `quarantine` — see BackendOutagePolicy's class doc — never merged with it.
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(System::nanoTime);
         PeerLauncher workers = new CompositePeerLauncher(
                 adapters,
                 cfg.effectiveDefaultProfile(),
                 config,
                 profileName -> liveCountRef.get().apply(profileName),
-                quarantine);
+                quarantine,
+                outagePolicy);
         // CB-504: under supervision (launchd/systemd) fleetd can start before herdr's socket
         // exists. The client itself is lazy — it connects per call — but the orphan reap below is
         // the first thing that actually talks to herdr, so without this wait a boot-order race
@@ -341,7 +352,29 @@ public final class Fleetd {
                 .map(session -> exhaustedPatternsByProfile.get(session.profile()))
                 .orElse(null);
         log.info("backend-exhausted classification (CB-578 stage A): {}",
-                CompletionResolver.coverage(cfg.profiles().keySet(), exhaustedPatternsByProfile.keySet()));
+                CompletionResolver.coverage("exhaustedPattern", cfg.profiles().keySet(),
+                        exhaustedPatternsByProfile.keySet()));
+        // fleetd #201 Unit 5: classify a completion-fallback scrape that matches a profile's
+        // configured backend-error refusal (a credential outage, a provider 5xx) as a backend error
+        // rather than handing it back as a real answer. Compiled once at startup, keyed by profile
+        // name, mirroring exhaustedPatternsByProfile above — a profile with no configured
+        // errorPattern is simply absent here, so CompletionResolver falls back to its built-in
+        // narrow {@code (?i)\bAPI Error\s*:} compatibility pattern for that profile's targets
+        // (BackendErrorPatternLookup#legacy's contract — see backendErrorPatterns below).
+        Map<String, Pattern> errorPatternsByProfile = new LinkedHashMap<>();
+        cfg.profiles().forEach((name, profile) -> {
+            if (profile.hasErrorPattern()) {
+                errorPatternsByProfile.put(name, Pattern.compile(profile.errorPattern()));
+            }
+        });
+        // fleetd #248: extracted to a static factory (see backendErrorPatternLookup below) so a
+        // test can prove main() actually PASSES this into CompletionResolver, not only that the
+        // lookup itself behaves correctly — the exact gap fleetd #248 exists to close.
+        BackendErrorPatternLookup backendErrorPatterns = backendErrorPatternLookup(sessions::roster,
+                errorPatternsByProfile);
+        log.info("backend-error classification (fleetd #201 Unit 5): {}",
+                CompletionResolver.coverage("errorPattern", cfg.profiles().keySet(),
+                        errorPatternsByProfile.keySet()));
         // CB-578 stage B: on a classification that actually wins, quarantine the exhausted profile's
         // CREDENTIAL — not the profile name — so a profile sharing that credential (e.g. two models
         // on one OpenAI account) is refused too, not just the one that happened to report it. Reads
@@ -389,13 +422,30 @@ public final class Fleetd {
         // fleetd #175: point the forwarding sink handed to OpenCodeLauncher above at the real one,
         // now that `sessions` exists to resolve target -> session -> profile.
         exhaustionSinkRef.set(exhaustionSink);
+        // fleetd #201 Unit 5: the production BackendErrorSink needs `pushLoop` (built further below,
+        // after `sessions`) to tell a lead about an incident or an unmapped target — the same
+        // construction-order cycle `exhaustionSinkRef` breaks above, broken the same way: a mutable
+        // holder set once `pushLoop` exists, read lazily from inside the lambda built here.
+        AtomicReference<ReplyPushLoop> pushLoopRef = new AtomicReference<>();
+        // fleetd #248: extracted to a static factory (see backendErrorSink below), public rather
+        // than package-private like the other two factories here, so
+        // dev.ltms.fleet.inject.BackendOutageFlowTest can exercise the REAL production sink
+        // directly instead of a hand-mirrored copy of this lambda — that copy was precisely the
+        // gap fleetd #248 exists to close (see that test's class doc for the history).
+        BackendErrorSink backendErrorSink = backendErrorSink(sessions, () -> config.get().profiles(),
+                outagePolicy, pushLoopRef::get);
         AgentControl agents = router.memberAgents();
-        CompletionResolver completion = new CompletionResolver(agents, rendezvous, exhaustedPatterns, exhaustionSink,
-                target -> sessions.roster().stream()
-                        .filter(session -> target.equals(session.terminalId()))
-                        .findFirst()
-                        .map(session -> new CompletionResolver.WorktreeBranch(session.worktree(), session.branch()))
-                        .orElse(null));
+        // Both fleetd#201 Unit 5 (backend-error patterns + sink) and fleetd#241 (the worktree/branch
+        // lookup the fallback report names) land on this one call. The full constructor takes both,
+        // so neither feature is dropped; nowNanos must be passed explicitly to reach it.
+        //
+        // fleetd #248: every argument built specifically for this call (backendErrorPatterns and
+        // backendErrorSink above, and the worktree/branch lookup right here) now comes from a
+        // static factory tested on its own; FleetdCompletionResolverWiringTest source-asserts that
+        // THIS call actually passes them, which is the coverage that was missing before.
+        CompletionResolver completion = new CompletionResolver(agents, rendezvous, exhaustedPatterns,
+                exhaustionSink, backendErrorPatterns, backendErrorSink, System::nanoTime,
+                worktreeBranchLookup(sessions::roster));
         // CB-113: deliver only to an available worker (its MCP is connected), never its boot window.
         // CB-301: the manager's presence bridge records availability and drives SPAWNING → READY.
         MemberPresence presence = sessions.asPresence();
@@ -477,6 +527,9 @@ public final class Fleetd {
         Metrics metrics = FleetMetrics.create(sessions, replyInbox);
         var pushLoop = new ReplyPushLoop(primaryRegistry, router.leadAgents(), replyInbox,
                 pushScheduler, maxReminders, backoffMs, metrics);
+        // fleetd #201 Unit 5: point the forwarding holder captured by the backendErrorSink lambda
+        // above at the real push loop, now that it exists.
+        pushLoopRef.set(pushLoop);
         // CB-551: idle-lead heartbeat. Opt-in; absent `leadHeartbeat:` this is never constructed, so
         // an upgraded daemon cannot silently start spending subscription on nudging an idle lead.
         // It has its own single-thread scheduler and holds its own scheduler shutdown via close().
@@ -584,7 +637,11 @@ public final class Fleetd {
                     var configured = config.get().profiles().get(profile);
                     return configured == null ? null : configured.effectiveCredentialId();
                 }, quarantine),
-                leadMailbox);
+                leadMailbox,
+                new FleetMcp.OutageSource(profile -> {
+                    var configured = config.get().profiles().get(profile);
+                    return configured == null ? null : configured.effectiveCredentialId();
+                }, outagePolicy));
 
         // CB-637: the receive half. Only constructed when a lead mailbox actually opened — with no
         // coordinator (or an unreachable one) there is nothing to deliver, so no scheduler is
@@ -652,8 +709,11 @@ public final class Fleetd {
 
         // CB-185: give FleetApp both daemons — /healthz must require both to answer and
         // GET /sessions must merge across both, or a down/unpolled member daemon is invisible.
+        // fleetd #111: live (re-read-per-request) memberCredentials view for GET /member-credentials —
+        // same hot-reload shape as the memberCredentials supplier passed to ClaudeCodeLauncher above.
         Javalin app = new FleetApp(herdr, memberHerdr, workers, sessions, messages, presence, mcp.servlet(),
-                callers, metrics, deliverable).build();
+                callers, metrics, deliverable,
+                () -> MemberCredentialPolicyView.of(config.get().memberCredentials())).build();
         app.start(cfg.bind().host(), cfg.bind().port());
         log.info("fleetd listening on {}:{}, herdr socket {}",
                 cfg.bind().host(), cfg.bind().port(), socket);
@@ -682,6 +742,123 @@ public final class Fleetd {
         return target -> presence.isPresent(target) || leads.get().containsKey(target);
     }
 
+    /**
+     * fleetd #248: package-private factory for the member worktree/branch lookup {@link
+     * CompletionResolver} uses to name a fallback report's worktree and branch (fleetd#241).
+     *
+     * <p>Before this ticket the lookup was an anonymous lambda built inline inside {@code main}'s
+     * {@code CompletionResolver} constructor call — provably untested wiring, the whole reason
+     * fleetd #248 exists: dropping that one argument (passing {@code _ -> null} instead) compiled
+     * clean and left every test green. Extracted here, {@code main} now calls this factory instead
+     * of building the lambda inline, and a source assertion on that call site
+     * ({@code FleetdCompletionResolverWiringTest}) proves the argument is still actually passed.
+     *
+     * <p>Takes the roster as a plain {@link Supplier} — not a {@link SessionManager} — so this is
+     * directly testable with a hand-built session list; no real {@code SessionManager} (launcher,
+     * worktrees, …) needs constructing. Follows the same {@code static} factory pattern as
+     * {@link #deliverableTo} above.
+     *
+     * @param roster the live member roster, normally {@code sessions::roster}
+     */
+    static Function<String, CompletionResolver.WorktreeBranch> worktreeBranchLookup(
+            Supplier<List<MemberSession>> roster) {
+        return target -> roster.get().stream()
+                .filter(session -> target.equals(session.terminalId()))
+                .findFirst()
+                .map(session -> new CompletionResolver.WorktreeBranch(session.worktree(), session.branch()))
+                .orElse(null);
+    }
+
+    /**
+     * fleetd #248 / fleetd#201 Unit 5: package-private factory for the per-target backend-error
+     * pattern lookup {@link CompletionResolver} classifies a pane scrape against. Closes over the
+     * live roster (to resolve a target to a profile) and {@code errorPatternsByProfile} (each
+     * profile's configured {@code errorPattern}, already compiled by the caller — the same map also
+     * feeds the coverage log next to where this is called) — nothing else, so it is directly
+     * testable. See {@link #worktreeBranchLookup} above for why this ticket exists and why the
+     * factory takes a roster {@link Supplier} rather than a {@link SessionManager}.
+     *
+     * @param roster                 the live member roster, normally {@code sessions::roster}
+     * @param errorPatternsByProfile every profile that has an {@code errorPattern} configured,
+     *                               keyed by profile name
+     */
+    static BackendErrorPatternLookup backendErrorPatternLookup(Supplier<List<MemberSession>> roster,
+            Map<String, Pattern> errorPatternsByProfile) {
+        return target -> roster.get().stream()
+                .filter(session -> target.equals(session.terminalId()))
+                .findFirst()
+                .map(session -> errorPatternsByProfile.get(session.profile()))
+                .orElse(null);
+    }
+
+    /**
+     * fleetd #248 / fleetd#201 Unit 5: factory for the production {@link BackendErrorSink} — the
+     * collaborator {@link CompletionResolver} notifies when a pane-scrape classification actually
+     * resolves a waiter as a backend error. Order: (1) mark the member BACKEND_ERROR; (2) resolve
+     * profile/credential through the roster — fail loud (never {@code Optional.ifPresent}, the
+     * fleetd #234 lesson applied to this sink) and notify the lead via {@code
+     * onBackendTargetUnmapped} when it cannot be resolved; (3) record the error in {@code
+     * outagePolicy}; (4) on a NEW incident (the {@code record()} call that actually crosses the
+     * threshold), tell the lead via {@code onBackendIncident}.
+     *
+     * <p>{@code public}, unlike {@link #worktreeBranchLookup} and {@link #backendErrorPatternLookup}
+     * above: {@code dev.ltms.fleet.inject.BackendOutageFlowTest} exercises this exact object as the
+     * real, wired production path, replacing what its own class doc used to call out as a
+     * hand-mirrored copy of this lambda ("mirrors {@code Fleetd.main}'s {@code backendErrorSink}
+     * lambda line-for-line") — that copy proved only itself, never that {@code main} still wires the
+     * real thing. That was precisely the gap fleetd #248 exists to close.
+     *
+     * @param sessions the session registry; both read (roster) and written (onBackendError)
+     * @param profiles the live profile map, normally {@code () -> config.get().profiles()} in
+     *                 {@code main}, or a fixed test map via {@code () -> profiles}
+     * @param pushLoop the lead-nudge loop, read lazily: {@code main} builds this sink before the
+     *                 real {@link ReplyPushLoop} exists (a genuine construction-order cycle, broken
+     *                 the same way {@code exhaustionSinkRef} is a few lines above it), so a
+     *                 {@link Supplier} reads whatever {@code main} has filled in by the time a real
+     *                 backend error fires
+     */
+    public static BackendErrorSink backendErrorSink(SessionManager sessions,
+            Supplier<Map<String, FleetConfig.Profile>> profiles, BackendOutagePolicy outagePolicy,
+            Supplier<ReplyPushLoop> pushLoop) {
+        return (target, matchedLine, reason) -> {
+            sessions.onBackendError(target, reason);
+
+            String profileName = sessions.roster().stream()
+                    .filter(session -> target.equals(session.terminalId()))
+                    .findFirst()
+                    .map(MemberSession::profile)
+                    .orElse(null);
+            FleetConfig.Profile profile = profileName == null ? null : profiles.get().get(profileName);
+            if (profile == null) {
+                log.error("backend error on target '{}' ({}) but no profile could be resolved — the "
+                                + "target is not (yet) in the roster — no cool-off applied (fleetd #201 Unit 5)",
+                        target, reason);
+                ReplyPushLoop loop = pushLoop.get();
+                if (loop != null) {
+                    loop.onBackendTargetUnmapped(target, reason);
+                }
+                return;
+            }
+            String credentialId = profile.effectiveCredentialId();
+            Optional<BackendOutagePolicy.Incident> incident = outagePolicy.record(credentialId, target, reason);
+            incident.ifPresent(inc -> {
+                List<String> affectedProfiles = profiles.get().values().stream()
+                        .filter(p -> credentialId.equals(p.effectiveCredentialId()))
+                        .map(FleetConfig.Profile::profile)
+                        .sorted()
+                        .toList();
+                log.warn("credential '{}' cooling off for {}s after backend errors on {} distinct "
+                                + "target(s) (profile '{}'): {}", credentialId,
+                        inc.remainingCoolOffSeconds(), inc.evidenceCount(), profile.profile(), reason);
+                ReplyPushLoop loop = pushLoop.get();
+                if (loop != null) {
+                    loop.onBackendIncident(inc.id(), inc.targets(), credentialId, affectedProfiles,
+                            (int) inc.remainingCoolOffSeconds());
+                }
+            });
+        };
+    }
+
     /** Injection seam for {@link #selectReplyInbox}: production binds {@link AmqpReplyInbox#open}. */
     @FunctionalInterface
     interface AmqpOpener {
@@ -909,12 +1086,14 @@ public final class Fleetd {
      * #requiredSecretEnvVars} is exposed for {@link #reportRequiredSecrets}'s own test.
      */
     static void reportMemberCredentialsGap(FleetConfig cfg) {
-        FleetConfig.MemberCredentials creds = cfg.memberCredentials();
-        if (creds != null && !creds.known().isEmpty()) {
+        // fleetd #111: the counts below come from MemberCredentialPolicyView, the same class the
+        // live GET /member-credentials endpoint reads — one place computes them, not two.
+        MemberCredentialPolicyView view = MemberCredentialPolicyView.of(cfg.memberCredentials());
+        if (view.present()) {
             log.info("memberCredentials: policy={}, {} known name(s), {} allowed — blocking {} on "
                             + "every spawn{}",
-                    creds.policy(), creds.known().size(), creds.allow().size(), creds.blockedSet().size(),
-                    creds.isAllowList()
+                    view.policy(), view.knownCount(), view.allowedCount(), view.blockedCount(),
+                    cfg.memberCredentials().isAllowList()
                             ? " (allow-list: known/allow are reporting only — the control is the derived ZDOTDIR scrub)"
                             : "");
             return;

fleetd/src/main/java/dev/ltms/fleet/config/ConfigRef.java

@@ -43,7 +43,9 @@ import java.util.function.Supplier;
  *       which is constructed once), <em>and an existing profile's launch settings</em> —
  *       {@code model}, {@code baseUrl}, {@code argv}, {@code env}, {@code mcpUrl},
  *       {@code exhaustedPattern} (CB-578 stage A — compiled once into {@code Fleetd.main}'s
- *       pattern map at startup), and the rest. {@code credentialId} (CB-578 stage B) is NOT on
+ *       pattern map at startup), {@code errorPattern} (fleetd #201 Unit 5 — compiled once into
+ *       {@code Fleetd.main}'s backend-error pattern map at startup, the same way), and the rest.
+ *       {@code credentialId} (CB-578 stage B) is NOT on
  *       this list — it is read live off the config supplier at every quarantine check and
  *       exhaustion event, exactly like {@code weight} / {@code maxLoad}, so it is hot instead.
  *       {@code HerdrPeerLauncher} takes {@code Map.copyOf(profiles)} at construction and resolves
@@ -292,6 +294,10 @@ public final class ConfigRef implements Supplier<FleetConfig> {
                 // CB-578 stage B: exhaustedPattern is compiled once into Fleetd.main's pattern map
                 // at startup (see ExhaustedPatternLookup wiring) — a reload never re-reads it, so a
                 // changed pattern must be reported as deferred, exactly like model/baseUrl/argv.
-                && Objects.equals(a.exhaustedPattern(), b.exhaustedPattern());
+                && Objects.equals(a.exhaustedPattern(), b.exhaustedPattern())
+                // fleetd #201 Unit 5: errorPattern is compiled once into Fleetd.main's backend-error
+                // pattern map at startup (see BackendErrorPatternLookup wiring), the same way
+                // exhaustedPattern is — a reload never re-reads it either.
+                && Objects.equals(a.errorPattern(), b.errorPattern());
     }
 }

fleetd/src/main/java/dev/ltms/fleet/config/FleetConfig.java

@@ -24,6 +24,8 @@ import java.util.LinkedHashMap;
 import java.util.List;
 import java.util.Map;
 import java.util.Set;
+import java.util.regex.Pattern;
+import java.util.regex.PatternSyntaxException;
 
 /**
  * {@code fleetd} configuration, loaded from a YAML file (see
@@ -257,7 +259,13 @@ public record FleetConfig(
      * @param cwd            fixed working directory for this profile's workers (CB-112 "told otherwise");
      *                       {@code null}/blank → inherit the primary's cwd, else the daemon's
      * @param parityOverlay  repo-relative paths copied primary→worktree for config parity; null/empty
-     *                       defaults to a sensible set of local config files.
+     *                       defaults to {@code [.env]} only (CB-148 point 2). {@code .env} is data — a
+     *                       copy of it can only carry values. {@code .envrc} is executable shell that
+     *                       {@code direnv} evaluates on every {@code cd}, so copying it moves behaviour
+     *                       into the worker, not just values, and that is a different risk from copying
+     *                       data. It is deliberately left out of the default for that reason. The knob
+     *                       still supports it: an operator who wants it copied writes
+     *                       {@code parityOverlay: [.env, .envrc]} explicitly and owns that choice.
      *                       <p><strong>Every overlay path must be gitignored or tracked-and-skipped.</strong>
      *                       CB-576 made {@code release()} preserve a worktree that {@code git status
      *                       --porcelain} reports as dirty, and it deliberately counts untracked files —
@@ -267,11 +275,10 @@ public record FleetConfig(
      *                       worktrees then accumulate with no error anywhere.
      *                       <p>Checked on 2026-08-15 (CB-581): inert as configured. Tracked overlay
      *                       files carry {@code --skip-worktree} so {@code --porcelain} cannot see them,
-     *                       {@code fleetd.yaml} is gitignored, and the default pair {@code .env} /
-     *                       {@code .envrc} does not exist in this repo. Note the default applies to
-     *                       <em>every</em> profile, so creating either file at the repo root is enough
-     *                       to make it live. Add a new overlay path to {@code .gitignore} in the same
-     *                       change that adds it here.
+     *                       {@code fleetd.yaml} is gitignored, and {@code .env} does not exist in this
+     *                       repo. Note the default applies to <em>every</em> profile, so creating
+     *                       {@code .env} at the repo root is enough to make it live. Add a new overlay
+     *                       path to {@code .gitignore} in the same change that adds it here.
      *                       <p>CB-578 stage C raised the stakes: a preserved dirty worktree is now also
      *                       committed to {@code refs/wip/<branch>} via {@code git add -A}. The overlay
      *                       carries the primary's own environment files into the worktree, so a
@@ -356,6 +363,16 @@ public record FleetConfig(
      *                       {@code limit.context} instead, which bounds the window opencode compacts
      *                       <em>within</em>, and only when {@code model} resolves to a
      *                       {@code provider/model} pair.
+     * @param errorPattern   regex matched against a completion-fallback scrape (fleetd #201 Unit 5)
+     *                       to classify a turn that ended with no {@code fleet_reply} as a backend
+     *                       error (credential outage, provider 5xx) rather than a real answer or an
+     *                       exhausted usage limit. {@code null}/blank ⇒ this profile relies on
+     *                       {@link dev.ltms.fleet.inject.CompletionResolver}'s built-in
+     *                       {@code (?i)\bAPI Error\s*:} compatibility pattern instead — classification
+     *                       still happens, just without a profile-specific match (reported separately
+     *                       at startup as legacy-default coverage, never full coverage). Compiled once
+     *                       at startup ({@code Fleetd.main}), like {@code exhaustedPattern}, so a
+     *                       reload of this key is DEFERRED (see {@code ConfigRef}).
      */
     @JsonIgnoreProperties(ignoreUnknown = true)
     public record Profile(String profile, String baseUrl, String model,
@@ -374,7 +391,8 @@ public record FleetConfig(
                          String ideMcpUrl,
                          String ideProjectDir,
                          String ideOpenCommand,
-                         Integer autoCompactWindow) {
+                         Integer autoCompactWindow,
+                         String errorPattern) {
 
         /** Peer kind spawned by {@link dev.ltms.fleet.member.ClaudeCodeLauncher} (the default). */
         public static final String KIND_CLAUDE_CODE = "claude-code";
@@ -408,7 +426,7 @@ public record FleetConfig(
             // servers do not exist there to be enabled. This is defence in depth, not a live fix: the
             // worker stays isolated only because this separate mechanism already removes the servers.
             parityOverlay = (parityOverlay == null || parityOverlay.isEmpty())
-                    ? List.of(".env", ".envrc")
+                    ? List.of(".env")
                     : List.copyOf(parityOverlay);
             // gitTokenEnv stays null when unset (opt-in). gitHostEnv defaults so operators enabling
             // checkpoints need only set gitTokenEnv; it is injected only alongside a resolved token.
@@ -448,6 +466,32 @@ public record FleetConfig(
             // there is no blank-string form to normalize (it's an Integer), and the [100000, 1000000]
             // range is enforced eagerly at config load (rejectAutoCompactWindowOutOfRange), naming the
             // profile, rather than silently clamped here. A profile that never sets it keeps null.
+            // errorPattern stays null when unset/blank (opt-in) — same rule as exhaustedPattern: no
+            // defaulting, no vendor wording. A profile that never sets it relies on
+            // CompletionResolver's built-in compatibility pattern instead (never "off").
+            errorPattern = (errorPattern == null || errorPattern.isBlank()) ? null : errorPattern;
+        }
+
+        /**
+         * Backward-compatible constructor without the fleetd #201 {@code errorPattern} field — the
+         * profile relies on {@code CompletionResolver}'s built-in compatibility pattern instead
+         * (legacy-default coverage). This is the shape the canonical constructor had before the
+         * field was added; every pre-existing Java call site (and any YAML that omits the key) keeps
+         * compiling and behaving identically. Jackson binds the canonical (longest) constructor, so
+         * YAML omitting {@code errorPattern:} still lands here as {@code null} via that path too.
+         */
+        public Profile(String profile, String baseUrl, String model,
+                      String configDir, String tokenEnv, List<String> argv,
+                      String placement, String workspace, String tabLabel, String mcpUrl,
+                      String cwd, List<String> parityOverlay, String gitTokenEnv, String gitHostEnv,
+                      String kind, Map<String, String> env, Float weight, Integer maxLoad,
+                      Boolean subscription, String exhaustedPattern, String credentialId,
+                      String ideMcpUrl, String ideProjectDir, String ideOpenCommand,
+                      Integer autoCompactWindow) {
+            this(profile, baseUrl, model, configDir, tokenEnv, argv, placement, workspace, tabLabel,
+                    mcpUrl, cwd, parityOverlay, gitTokenEnv, gitHostEnv, kind, env, weight, maxLoad,
+                    subscription, exhaustedPattern, credentialId, ideMcpUrl, ideProjectDir, ideOpenCommand,
+                    autoCompactWindow, null);
         }
 
         /**
@@ -492,7 +536,8 @@ public record FleetConfig(
         public Profile withProfile(String p) {
             return new Profile(p, baseUrl, model, configDir, tokenEnv, argv, placement, workspace, tabLabel,
                     mcpUrl, cwd, parityOverlay, gitTokenEnv, gitHostEnv, kind, env, weight, maxLoad, subscription,
-                    exhaustedPattern, credentialId, ideMcpUrl, ideProjectDir, ideOpenCommand, autoCompactWindow);
+                    exhaustedPattern, credentialId, ideMcpUrl, ideProjectDir, ideOpenCommand, autoCompactWindow,
+                    errorPattern);
         }
 
         /**
@@ -589,6 +634,15 @@ public record FleetConfig(
             return exhaustedPattern != null;
         }
 
+        /**
+         * True when this profile configures its own fleetd #201 Unit 5 backend-error classification
+         * pattern. {@code false} means this profile relies on {@code CompletionResolver}'s built-in
+         * {@code (?i)\bAPI Error\s*:} compatibility pattern instead (legacy-default coverage).
+         */
+        public boolean hasErrorPattern() {
+            return errorPattern != null;
+        }
+
         /**
          * The credential group this profile quarantines with (CB-578 stage B): the configured
          * {@link #credentialId} when set, else this profile's own name — so an unconfigured profile
@@ -1389,6 +1443,7 @@ public record FleetConfig(
             rejectDuplicateMemberSlots(yaml);
             rejectNegativeMaxLoad(yaml);
             rejectAutoCompactWindowOutOfRange(yaml);
+            rejectMalformedErrorPattern(yaml);
             rejectUnknownKind(yaml);
             rejectUnknownAuthMode(yaml);
             rejectUnknownPlacement(yaml);
@@ -1737,6 +1792,51 @@ public record FleetConfig(
         }
     }
 
+    /**
+     * Reject a profile whose {@code errorPattern} (fleetd #201 Unit 5) is not a valid Java regex,
+     * naming the profile, the key, and the parser's own message.
+     *
+     * <p>Unset/{@code null} means "use {@code CompletionResolver}'s built-in {@code (?i)\bAPI
+     * Error\s*:} compatibility pattern" and passes silently. A profile that DOES set the key gets it
+     * compiled once at daemon startup ({@code Fleetd.main}, mirroring {@code exhaustedPattern}) — an
+     * uncaught {@link java.util.regex.PatternSyntaxException} there crashes startup without naming
+     * which profile or key is at fault. Validate eagerly here instead, at config load, the same
+     * "fail loud at load, not lazily later" reasoning as {@link #rejectAutoCompactWindowOutOfRange}.
+     *
+     * @param yaml the raw config text
+     * @throws IllegalStateException when any profile's {@code errorPattern} fails to compile
+     */
+    static void rejectMalformedErrorPattern(String yaml) {
+        Map<?, ?> raw;
+        try {
+            raw = YAML.readValue(yaml, Map.class);
+        } catch (IOException | IllegalArgumentException e) {
+            return; // a malformed file is reported by the real parse, not here
+        }
+        if (raw == null || !(raw.get("profiles") instanceof Map<?, ?> profiles)) {
+            return;
+        }
+        List<String> bad = new ArrayList<>();
+        for (Map.Entry<?, ?> e : profiles.entrySet()) {
+            if (!(e.getValue() instanceof Map<?, ?> p)) {
+                continue;
+            }
+            if (!(p.get("errorPattern") instanceof String pattern) || pattern.isBlank()) {
+                continue;
+            }
+            try {
+                Pattern.compile(pattern);
+            } catch (PatternSyntaxException ex) {
+                bad.add("profiles." + e.getKey() + ".errorPattern (\"" + pattern + "\"): " + ex.getMessage());
+            }
+        }
+        bad.sort(String::compareTo);
+        if (!bad.isEmpty()) {
+            throw new IllegalStateException("refusing to start: malformed errorPattern — "
+                    + String.join("; ", bad));
+        }
+    }
+
     /** The peer kinds this build has an adapter for — {@link Profile#kind()}'s only valid values. */
     private static final Set<String> KNOWN_KINDS = Set.of(Profile.KIND_CLAUDE_CODE, Profile.KIND_OPENCODE);
 

fleetd/src/main/java/dev/ltms/fleet/inject/CompletionResolver.java

@@ -620,9 +620,9 @@ public final class CompletionResolver implements TurnListener {
      * @param allProfiles        every configured profile name
      * @param configuredProfiles the subset of {@code allProfiles} that carry an exhausted pattern
      */
-    public static String coverage(Set<String> allProfiles, Set<String> configuredProfiles) {
+    public static String coverage(String patternKey, Set<String> allProfiles, Set<String> configuredProfiles) {
         if (configuredProfiles.isEmpty()) {
-            return "off (no profile has an exhaustedPattern configured; profiles: " + sorted(allProfiles) + ")";
+            return "off (no profile has an " + patternKey + " configured; profiles: " + sorted(allProfiles) + ")";
         }
         Set<String> unconfigured = new TreeSet<>(allProfiles);
         unconfigured.removeAll(configuredProfiles);

fleetd/src/main/java/dev/ltms/fleet/mcp/FleetMcp.java

@@ -17,6 +17,7 @@ import dev.ltms.fleet.msg.LeadMessage;
 import dev.ltms.fleet.msg.MessageService;
 import dev.ltms.fleet.msg.Rendezvous;
 import dev.ltms.fleet.peer.PeerUnreachableException;
+import dev.ltms.fleet.placement.BackendOutagePolicy;
 import dev.ltms.fleet.placement.BackendQuarantine;
 import dev.ltms.fleet.placement.PlacementException;
 import dev.ltms.fleet.session.SessionManager;
@@ -93,6 +94,8 @@ public final class FleetMcp {
     private final CapacitySource capacity;
     private final HealthCoverageSource healthCoverage;
     private final QuarantineSource quarantine;
+    /** fleetd #201 Unit 5: SEPARATE from {@link #quarantine} — see {@link OutageSource}'s doc. */
+    private final OutageSource outage;
     /** CB-637: this daemon's lead-to-lead channel; {@code null} when no coordinator is configured. */
     private final LeadChannel leadChannel;
 
@@ -116,6 +119,22 @@ public final class FleetMcp {
         public static QuarantineSource none() { return new QuarantineSource(_ -> null, BackendQuarantine.none()); }
     }
 
+    /**
+     * fleetd #201 Unit 5 cool-off facts used by {@code fleet_list}/{@code fleet_profiles}: a
+     * profile → credential id lookup, plus the shared {@link BackendOutagePolicy} to read remaining
+     * cool-offs off. A SEPARATE source from {@link QuarantineSource} — never merged into it — so a
+     * credential that is cooling off after repeated backend errors is reported independently of
+     * whether it is ALSO under CB-578 stage B exhaustion quarantine; the two checks can both fire
+     * for the same profile, and when they do, {@code fleet_list}/{@code fleet_profiles} report both
+     * (see {@link #capacityView}/{@link #profiles}).
+     */
+    public record OutageSource(Function<String, String> credentialIdFor, BackendOutagePolicy outagePolicy) {
+        /** Inert source — no profile is ever reported cooling off. Explicit stand-in, not a default. */
+        public static OutageSource none() {
+            return new OutageSource(_ -> null, new BackendOutagePolicy(System::nanoTime));
+        }
+    }
+
     /**
      * @param callers resolves each call's {@link Principal}; {@code null} disables authorization.
      *                This surface needs its own enforcement: {@code /mcp} is a raw servlet on
@@ -130,7 +149,7 @@ public final class FleetMcp {
                       CallerResolver callers, Metrics metrics, CapacitySource capacity, HealthCoverageSource healthCoverage,
                       QuarantineSource quarantine) {
         this(messages, workers, sessions, identity, presence, primaryRegistry, callers, metrics, capacity,
-                healthCoverage, quarantine, null);
+                healthCoverage, quarantine, null, OutageSource.none());
     }
 
     /**
@@ -143,9 +162,25 @@ public final class FleetMcp {
                      ConnectionIdentity identity, MemberPresence presence, PrimaryRegistry primaryRegistry,
                       CallerResolver callers, Metrics metrics, CapacitySource capacity, HealthCoverageSource healthCoverage,
                       QuarantineSource quarantine, LeadChannel leadChannel) {
+        this(messages, workers, sessions, identity, presence, primaryRegistry, callers, metrics, capacity,
+                healthCoverage, quarantine, leadChannel, OutageSource.none());
+    }
+
+    /**
+     * As above, with fleetd #201 Unit 5 cool-off facts for {@code fleet_list}/{@code fleet_profiles}
+     * (see {@link OutageSource}). This is what {@code Fleetd.main} actually wires up.
+     *
+     * @param outage required — pass {@link OutageSource#none()} for a caller that does not want the
+     *               feature, never a defaulting overload (the same rule {@code quarantine} follows).
+     */
+    public FleetMcp(MessageService messages, PeerLauncher workers, SessionManager sessions,
+                     ConnectionIdentity identity, MemberPresence presence, PrimaryRegistry primaryRegistry,
+                      CallerResolver callers, Metrics metrics, CapacitySource capacity, HealthCoverageSource healthCoverage,
+                      QuarantineSource quarantine, LeadChannel leadChannel, OutageSource outage) {
         this.leadChannel = leadChannel;
         this.capacity = capacity;
         this.quarantine = Objects.requireNonNull(quarantine, "quarantine");
+        this.outage = Objects.requireNonNull(outage, "outage");
         this.healthCoverage = healthCoverage;
         McpJsonMapper json = new JacksonMcpJsonMapperSupplier().get();
         this.transport = HttpServletStreamableServerTransportProvider.builder()
@@ -274,7 +309,7 @@ public final class FleetMcp {
                 (exchange, _) -> {
                     McpSchema.CallToolResult denied = deny(exchange, Authz.Action.READ, null);
                     if (denied != null) return denied;
-                    return listFleet(workers, sessions, messages, capacity, healthCoverage, quarantine,
+                    return listFleet(workers, sessions, messages, capacity, healthCoverage, quarantine, outage,
                             callers == null ? Map.of() : callers.leads(),
                             callerTerminal(exchange),
                             leadChannel == null ? null : leadChannel.selfCoordId());
@@ -290,7 +325,7 @@ public final class FleetMcp {
                 (exchange, _) -> {
                     McpSchema.CallToolResult denied = deny(exchange, Authz.Action.READ, null);
                     if (denied != null) return denied;
-                    return profiles(workers, quarantine);
+                    return profiles(workers, quarantine, outage);
                 };
         BiFunction<McpSyncServerExchange, McpSchema.CallToolRequest, McpSchema.CallToolResult> whoamiHandler =
                 (exchange, _) -> {
@@ -875,25 +910,47 @@ public final class FleetMcp {
      * has ever been quarantined gets exactly the pre-stage-B shape.
      */
     static McpSchema.CallToolResult profiles(PeerLauncher workers, QuarantineSource quarantine) {
+        return profiles(workers, quarantine, OutageSource.none());
+    }
+
+    /**
+     * As above, plus a SEPARATE {@code coolingOff} map (fleetd #201 Unit 5) — never merged into
+     * {@code quarantined} — for a profile whose credential is cooling off after repeated backend
+     * errors ({@link BackendOutagePolicy}). The two checks are independent: a profile can appear in
+     * both maps at once when it is both exhaustion-quarantined AND cooling off.
+     */
+    static McpSchema.CallToolResult profiles(PeerLauncher workers, QuarantineSource quarantine, OutageSource outage) {
         Map<String, Object> result = new LinkedHashMap<>();
         result.put("profiles", workers.profiles());
         result.put("default", workers.defaultProfile() == null ? "" : workers.defaultProfile());
         Map<String, Object> quarantined = new LinkedHashMap<>();
+        Map<String, Object> coolingOff = new LinkedHashMap<>();
         for (String profile : workers.profiles()) {
             String credentialId = quarantine.credentialIdFor().apply(profile);
-            if (credentialId == null) {
-                continue;
+            if (credentialId != null) {
+                quarantine.quarantine().remainingSeconds(credentialId).ifPresent(remaining -> {
+                    Map<String, Object> row = new LinkedHashMap<>();
+                    row.put("credentialId", credentialId);
+                    row.put("quarantinedForSeconds", remaining);
+                    quarantined.put(profile, row);
+                });
+            }
+            String outageCredentialId = outage.credentialIdFor().apply(profile);
+            if (outageCredentialId != null) {
+                outage.outagePolicy().remainingCoolOffSeconds(outageCredentialId).ifPresent(remaining -> {
+                    Map<String, Object> row = new LinkedHashMap<>();
+                    row.put("credentialId", outageCredentialId);
+                    row.put("coolingOffForSeconds", remaining);
+                    coolingOff.put(profile, row);
+                });
             }
-            quarantine.quarantine().remainingSeconds(credentialId).ifPresent(remaining -> {
-                Map<String, Object> row = new LinkedHashMap<>();
-                row.put("credentialId", credentialId);
-                row.put("quarantinedForSeconds", remaining);
-                quarantined.put(profile, row);
-            });
         }
         if (!quarantined.isEmpty()) {
             result.put("quarantined", quarantined);
         }
+        if (!coolingOff.isEmpty()) {
+            result.put("coolingOff", coolingOff);
+        }
         return text(json(result));
     }
 
@@ -932,6 +989,14 @@ public final class FleetMcp {
         return listFleet(workers, sessions, messages, capacity, healthCoverage, quarantine, leads, selfTerm, null);
     }
 
+    /** As above, plus fleetd #201 Unit 5 cool-off facts (see {@link OutageSource}). */
+    static McpSchema.CallToolResult listFleet(PeerLauncher workers, SessionManager sessions, MessageService messages,
+                                               CapacitySource capacity, HealthCoverageSource healthCoverage,
+                                               QuarantineSource quarantine, OutageSource outage,
+                                               Map<String, String> leads, String selfTerm) {
+        return listFleet(workers, sessions, messages, capacity, healthCoverage, quarantine, outage, leads, selfTerm, null);
+    }
+
     /**
      * As above, additionally reporting this daemon's own lead coordination id (CB-637) when one is
      * configured and its channel opened. There is no peer-discovery surface yet — a lead addresses a
@@ -945,6 +1010,15 @@ public final class FleetMcp {
                                                CapacitySource capacity, HealthCoverageSource healthCoverage,
                                                QuarantineSource quarantine, Map<String, String> leads, String selfTerm,
                                                String selfCoordId) {
+        return listFleet(workers, sessions, messages, capacity, healthCoverage, quarantine, OutageSource.none(),
+                leads, selfTerm, selfCoordId);
+    }
+
+    /** As above, plus fleetd #201 Unit 5 cool-off facts (see {@link OutageSource}). */
+    static McpSchema.CallToolResult listFleet(PeerLauncher workers, SessionManager sessions, MessageService messages,
+                                               CapacitySource capacity, HealthCoverageSource healthCoverage,
+                                               QuarantineSource quarantine, OutageSource outage,
+                                               Map<String, String> leads, String selfTerm, String selfCoordId) {
         try {
             Map<String, Agent> live = workers.list().stream()
                     .map(Agent.class::cast)
@@ -971,7 +1045,7 @@ public final class FleetMcp {
             }
             if (capacity.available()) result.put("capacity", profiles.stream()
                     .map(profile -> capacityView(profile, capacity.liveCount(), capacity.maxLoad(), roster, messages,
-                            capacity.clock().getAsLong(), quarantine)).toList());
+                            capacity.clock().getAsLong(), quarantine, outage)).toList());
             return text(json(result));
         } catch (HerdrException e) {
             return error("herdr error listing the fleet: " + e.getMessage());
@@ -1004,10 +1078,17 @@ public final class FleetMcp {
      * reports, reusing {@link QuarantineSource} rather than a second lookup. Both new keys are
      * added only when the profile is actually quarantined, so an ordinary fleet's rows are
      * byte-identical to before this change.
+     *
+     * <p>fleetd #201 Unit 5: a profile whose credential is cooling off (a SEPARATE, independent
+     * check from quarantine — see {@link OutageSource}) also forces {@code free} to 0 and carries
+     * {@code credentialId}/{@code coolingOffForSeconds}, but never {@code quarantinedForSeconds} —
+     * that key is added only when exhaustion quarantine is ALSO active for this profile, since the
+     * two checks are independent and either, both, or neither can be true.
      */
     private static Map<String, Object> capacityView(String profile, Function<String, Integer> liveCount,
                                                      Function<String, Integer> maxLoad, List<MemberSession> roster,
-                                                     MessageService messages, long nowNanos, QuarantineSource quarantine) {
+                                                     MessageService messages, long nowNanos, QuarantineSource quarantine,
+                                                     OutageSource outage) {
         Integer cap = maxLoad.apply(profile);
         int live = liveCount.apply(profile);
         int reclaimable = (int) roster.stream().filter(s -> profile.equals(s.profile()))
@@ -1025,6 +1106,14 @@ public final class FleetMcp {
                 row.put("quarantinedForSeconds", remaining);
             });
         }
+        String outageCredentialId = outage.credentialIdFor().apply(profile);
+        if (outageCredentialId != null) {
+            outage.outagePolicy().remainingCoolOffSeconds(outageCredentialId).ifPresent(remaining -> {
+                row.put("free", 0);
+                row.put("credentialId", outageCredentialId);
+                row.put("coolingOffForSeconds", remaining);
+            });
+        }
         return row;
     }
 
@@ -1164,7 +1253,11 @@ public final class FleetMcp {
                         + "worktree:<ticket-slug> to provision an isolated git worktree. Pass resumeSessionId "
                         + "to relaunch onto a prior conversation instead of starting cold — this requires an "
                         + "explicit profile whose backend supports it (fleet_list shows agentSessionId for "
-                        + "resumable members), and is refused otherwise rather than silently starting fresh. "
+                        + "resumable members; it is absent for a member fleetd cannot reliably re-identify, "
+                        + "e.g. an opencode member spawned without a worktree), and is refused otherwise "
+                        + "rather than silently starting fresh. For an opencode profile, resumeSessionId "
+                        + "itself also requires worktree:true/<slug> on THIS spawn — without one fleetd can "
+                        + "never re-verify which conversation it actually resumed (fleetd #249). "
                         + "sessionName gives the member a display name in its own UI when the backend supports "
                         + "one. Returns the member's sessionId (use with fleet_send) and paneId (use with "
                         + "fleet_stop).",
@@ -1175,7 +1268,7 @@ public final class FleetMcp {
                         "worktree", Map.of("type", "string", "description", "'true' or a ticket slug — requests an isolated git worktree"),
                         "ticket", stringProp("Ticket slug when worktree:true"),
                         "sessionName", stringProp("Logical display name for the member's own session, when its backend supports one"),
-                        "resumeSessionId", stringProp("A prior member's agentSessionId (from fleet_list) to resume — requires an explicit profile that supports it")),
+                        "resumeSessionId", stringProp("A prior member's agentSessionId (from fleet_list) to resume — requires an explicit profile that supports it, and (for opencode) a worktree on this spawn too")),
                         List.of()));
     }
 
@@ -1196,9 +1289,14 @@ public final class FleetMcp {
                         + "discover a peer lead without being told its address. 'members' are the "
                         + "sessions delegated to — each with sessionId, paneId, role (architect/dev/"
                         + "reviewer), profile (the backend it runs on), state, optional "
-                        + "worktree/branch/owner/agentSessionId (the id to pass as fleet_spawn's "
-                        + "resumeSessionId to relaunch onto that same conversation, when the backend "
-                        + "supports it), and live herdr status. An empty 'members' "
+                        + "worktree/branch/owner/agentSessionId, and live herdr status. agentSessionId, "
+                        + "when present, is the id to pass as fleet_spawn's resumeSessionId to relaunch "
+                        + "onto that same conversation. It is ABSENT — not a guess — for a member fleetd "
+                        + "cannot reliably re-identify: some backends (e.g. opencode) resolve it from the "
+                        + "member's working directory, which only uniquely identifies a member when it "
+                        + "was spawned into its own fleetd-provisioned worktree (worktree:true/<slug>); a "
+                        + "member spawned without one shares its directory with others and never reports "
+                        + "an id, however long it runs (fleetd #249). An empty 'members' "
                         + "means no members are spawned; it says nothing about peers. When capacity "
                         + "facts are configured, a 'capacity' row per profile also reports free: 0 for "
                         + "a quarantined profile's credential (see fleet_profiles), whatever its "

fleetd/src/main/java/dev/ltms/fleet/member/ClaudeCodeLauncher.java

@@ -1,5 +1,8 @@
 package dev.ltms.fleet.member;
 
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import com.fasterxml.jackson.databind.node.ObjectNode;
 import dev.ltms.fleet.config.FleetConfig;
 import dev.ltms.fleet.guard.SubscriptionGuard;
 import dev.ltms.fleet.herdr.Agent;
@@ -14,6 +17,8 @@ import java.io.IOException;
 import java.io.UncheckedIOException;
 import java.nio.file.Files;
 import java.nio.file.Path;
+import java.nio.file.StandardCopyOption;
+import java.nio.file.attribute.PosixFileAttributeView;
 import java.util.EnumSet;
 import java.util.List;
 import java.util.Map;
@@ -47,6 +52,21 @@ public final class ClaudeCodeLauncher extends HerdrPeerLauncher {
 
     private static final Logger log = LoggerFactory.getLogger(ClaudeCodeLauncher.class);
 
+    /** JSON codec for the additive workspace-trust seed (fleetd #149) — Jackson's default settings. */
+    private static final ObjectMapper TRUST_JSON = new ObjectMapper();
+
+    /**
+     * Serialises every {@link #seedTrustDialog} read-modify-write for the whole daemon process.
+     * Two claude-code spawns starting at once are normal (fleetd runs several members in parallel
+     * routinely) and both would otherwise read the same {@code .claude.json}, add their own entry
+     * to their own in-memory copy, and write — the second write wins and the first spawn's trust
+     * entry silently disappears. A single process-wide lock is enough because every spawn on this
+     * daemon runs in this one JVM; it does not protect against a second daemon process or the
+     * operator's own Claude Code process writing at the same instant, which {@link #writeAtomically}
+     * covers instead (each writer only ever sees a fully-old or fully-new file, never a torn one).
+     */
+    private static final Object TRUST_JSON_LOCK = new Object();
+
     private final SubscriptionGuard guard;
 
     /**
@@ -263,6 +283,10 @@ public final class ClaudeCodeLauncher extends HerdrPeerLauncher {
         putIfPresent(workerEnv, "ANTHROPIC_MODEL", cfg.model());
         putIfPresent(workerEnv, "CLAUDE_CONFIG_DIR", cfg.configDir());
         applyGitToken(workerEnv, cfg);
+        // fleetd #149: seed the workspace-trust entry BEFORE this spawn ever reaches herdr — see
+        // seedTrustDialog for why, and isProvisionedWorktree for why this is gated to a worktree
+        // fleetd itself provisioned (never a real checkout, never an un-configured fallback cwd).
+        seedTrustDialog(cfg.configDir(), spec.cwd());
 
         // CB-547a: Claude Code can MINT its own session id, so fleetd chooses it — a fresh spawn
         // gets a UUID we pass as --session-id and return from agentSessionId(), so the resume
@@ -412,12 +436,12 @@ public final class ClaudeCodeLauncher extends HerdrPeerLauncher {
      */
     private static void writeIdeOverlay(String cwd, String projectPath) {
         try {
-            Path dotGit = Path.of(cwd, ".git");
-            if (!Files.isRegularFile(dotGit)) {
+            if (!isProvisionedWorktree(cwd)) {
                 // Not a provisioned worktree (primary's real checkout has a .git directory, or the
                 // cwd is not a repo at all). Never write into it.
                 return;
             }
+            Path dotGit = Path.of(cwd, ".git");
             // The overlay FILE lives at the worktree root (claude-code's cwd), but its CONTENT pins
             // project_path to the module dir the IDE opened (projectPath), not the worktree root.
             Files.writeString(Path.of(cwd, "CLAUDE.local.md"), PeerLauncher.ideOverlayText(projectPath));
@@ -447,6 +471,173 @@ public final class ClaudeCodeLauncher extends HerdrPeerLauncher {
         }
     }
 
+    /**
+     * fleetd #149: pre-seed the workspace-trust entry for {@code cwd} in Claude Code's own config
+     * file, BEFORE this launch ever reaches herdr (called from {@link #buildLaunch}, which always
+     * runs before the base class starts the process). Claude Code asks an interactive, un-timed
+     * "Is this a project you created or one you trust?" the first time it starts in a directory it
+     * has not seen before, and every {@code worktree: true} spawn lands in a brand-new directory —
+     * so without this seed the member sits on that dialog forever, never mounts the bridge MCP, and
+     * never calls {@code fleet_reply}. herdr reports it as healthy the whole time ({@code
+     * agent_status: blocked}, {@code interactive_ready: true}), so nothing else catches it. Measured
+     * live on fleet01 2026-08-23: an unseeded fresh cwd sat on the dialog indefinitely; a seeded one
+     * reached {@code idle} clean.
+     *
+     * <p>This is not a new grant — the operator already trusted this repo by configuring the
+     * profile against it, and a worktree is a checkout of that same repo.
+     *
+     * <p>The key Claude Code reads is per-project, in {@code .claude.json}: {@code
+     * projects.<cwd>.hasTrustDialogAccepted}. The file lives at {@code <configDir>/.claude.json}
+     * when the profile sets {@code CLAUDE_CONFIG_DIR} (mirrors this launcher's own env var above),
+     * else the default {@code ~/.claude.json} — the same file Claude Code itself would read either
+     * way, so this seeds exactly what the spawned peer is about to open.
+     *
+     * <p><b>Additive, not a rewrite.</b> {@code .claude.json} is large (tens of KB, dozens of
+     * projects) and Claude Code itself rewrites it while running, so this reads the file as a JSON
+     * tree (missing or unreadable → treated as an empty object) and changes only
+     * {@code projects.<cwd>.hasTrustDialogAccepted} (that key alone — see fleetd #247) —
+     * every other top-level key and every other project entry is written back untouched. Only the
+     * one project entry for {@code cwd} is replaced/created; an existing entry for a DIFFERENT cwd
+     * (or the operator's own project history) is never touched.
+     *
+     * <p>Best-effort, like {@link #writeIdeOverlay}: a failure here (unwritable configDir, a
+     * corrupt existing file, …) must never fail the spawn — it is logged at debug and swallowed. A
+     * peer that starts without the seed still starts; it just may hit the dialog fleetd #149
+     * describes.
+     *
+     * <p><b>Gated to a provisioned worktree</b> ({@link HerdrPeerLauncher#isProvisionedWorktree}) — see that
+     * method's javadoc for the incident that made this gate mandatory, not optional: this must
+     * never run against a real checkout or an un-configured fallback cwd, only the exact
+     * always-fresh-directory population fleetd #149 describes.
+     *
+     * <p><b>Atomic and lock-protected.</b> {@code .claude.json} is a live file — Claude Code itself
+     * rewrites it while running, and this daemon routinely spawns several members at once, each
+     * calling this method for its own cwd. Every write goes through {@link #writeAtomically} (a
+     * sibling-temp-file + {@code ATOMIC_MOVE}, never a truncate-in-place) so a crash mid-write or a
+     * concurrent reader never observes a half-written file, and through {@link #TRUST_JSON_LOCK} so
+     * two concurrent spawns' entries both survive instead of the second write silently discarding
+     * the first. Both exist because of a real incident: see {@link HerdrPeerLauncher#isProvisionedWorktree}'s javadoc
+     * and {@link #writeAtomically}'s javadoc.
+     *
+     * @param configDir the profile's {@code CLAUDE_CONFIG_DIR} ({@code cfg.configDir()}), or
+     *                  {@code null}/blank to target the default {@code ~/.claude.json}
+     * @param cwd       the spawn's resolved working directory — the exact key Claude Code will look
+     *                  up for itself once it starts there
+     */
+    private static void seedTrustDialog(String configDir, String cwd) {
+        if (!isProvisionedWorktree(cwd)) {
+            return;
+        }
+        Path target = (configDir == null || configDir.isBlank())
+                ? Path.of(System.getProperty("user.home"), ".claude.json")
+                : Path.of(configDir, ".claude.json");
+        synchronized (TRUST_JSON_LOCK) {
+            try {
+                if (target.getParent() != null) {
+                    Files.createDirectories(target.getParent());
+                }
+                ObjectNode root = null;
+                if (Files.isRegularFile(target)) {
+                    JsonNode existing = TRUST_JSON.readTree(target.toFile());
+                    if (existing instanceof ObjectNode existingObject) {
+                        root = existingObject;
+                    }
+                }
+                if (root == null) {
+                    root = TRUST_JSON.createObjectNode();
+                }
+                JsonNode projectsNode = root.get("projects");
+                ObjectNode projects = projectsNode instanceof ObjectNode projectsObject
+                        ? projectsObject : TRUST_JSON.createObjectNode();
+                if (!(projectsNode instanceof ObjectNode)) {
+                    root.set("projects", projects);
+                }
+                JsonNode projectNode = projects.get(cwd);
+                ObjectNode project = projectNode instanceof ObjectNode projectObject
+                        ? projectObject : TRUST_JSON.createObjectNode();
+                if (!(projectNode instanceof ObjectNode)) {
+                    projects.set(cwd, project);
+                }
+                // fleetd #247: ONLY hasTrustDialogAccepted. We used to write
+                // hasCompletedProjectOnboarding beside it; do not put it back. Measured on
+                // 2026-09-03, minutes after a live spawn seeded this file: 28 of 28 project
+                // entries carried hasTrustDialogAccepted and 0 of 28 carried the onboarding key
+                // — including the 27 entries Claude Code wrote for itself. Claude Code
+                // normalises the whole file when it saves and drops that key every time, so
+                // writing it achieved nothing except making the next reader think it mattered.
+                // The member reached idle with the trust flag alone, which is the only outcome
+                // this seed exists for. If a future Claude Code needs the second flag the
+                // symptom returns as the trust dialog fleetd #149 describes — re-measure then,
+                // do not restore it on a guess.
+                project.put("hasTrustDialogAccepted", true);
+                writeAtomically(target, TRUST_JSON.writerWithDefaultPrettyPrinter().writeValueAsString(root));
+            } catch (Exception e) {
+                log.debug("cannot seed workspace-trust entry for cwd '{}' into '{}'", cwd, target, e);
+            }
+        }
+    }
+
+    /**
+     * Write {@code content} to {@code target} atomically: serialise to a sibling temp file in the
+     * <strong>same directory</strong> as {@code target} (an atomic move is only guaranteed within
+     * one filesystem — a different directory could mean a different filesystem), then
+     * {@link StandardCopyOption#ATOMIC_MOVE} it into place. A reader — Claude Code itself, or
+     * another {@code seedTrustDialog} call — only ever observes the fully-old file or the
+     * fully-new one, never a truncated or half-written one.
+     *
+     * <p><b>fleetd #149 incident.</b> The original implementation used
+     * {@code Files.writeString(target, content)} directly, which truncates {@code target} in place
+     * before writing the replacement bytes. Combined with an ungated {@code cwd} (see
+     * {@link HerdrPeerLauncher#isProvisionedWorktree}'s javadoc), a mutation-testing run hit that truncation window
+     * against the operator's real {@code ~/.claude.json} and left it at 178 bytes. The gate closes
+     * <em>which file</em> this can ever target; this closes <em>how</em> the target is written, so
+     * that even a legitimate write against a real, live, concurrently-read {@code .claude.json}
+     * cannot leave it observably empty or partial.
+     *
+     * <p>Preserves {@code target}'s existing POSIX permissions (Claude Code ships {@code
+     * .claude.json} as {@code 0600}) when the filesystem reports them; a freshly created temp file
+     * already defaults to owner-only permissions on a POSIX filesystem, so a first-ever write (no
+     * existing {@code target}) is no less private without this. On a non-POSIX filesystem (e.g.
+     * Windows) the permission copy is a silent no-op rather than a failure.
+     *
+     * <p>Package-visible (not {@code private}) so a test can drive it directly with a concurrent
+     * reader thread and prove the torn-file property this method exists for — the LOCK in
+     * {@link #seedTrustDialog} already fully serialises every call this launcher itself makes, so a
+     * test that only ever goes through {@code seedTrustDialog}/{@code spawn()} could never observe
+     * a torn file regardless of whether this method is atomic; it would be proving the lock, not
+     * this method. Atomicity's actual job is protecting against a writer the lock cannot reach at
+     * all — a second daemon process, or the operator's own live Claude Code — so the test for it
+     * has to reach this method on its own.
+     */
+    static void writeAtomically(Path target, String content) throws IOException {
+        Path parent = target.getParent();
+        Path tmp = Files.createTempFile(parent, target.getFileName() + ".", ".tmp");
+        try {
+            Files.writeString(tmp, content);
+            copyPosixPermissionsIfPresent(target, tmp);
+            Files.move(tmp, target, StandardCopyOption.ATOMIC_MOVE, StandardCopyOption.REPLACE_EXISTING);
+        } catch (IOException e) {
+            Files.deleteIfExists(tmp);
+            throw e;
+        }
+    }
+
+    /** Copy {@code target}'s POSIX permissions onto {@code tmp}, or no-op where either is unsupported. */
+    private static void copyPosixPermissionsIfPresent(Path target, Path tmp) {
+        try {
+            if (!Files.isRegularFile(target)) {
+                return; // nothing to inherit from — first-ever write, temp file's own default stands
+            }
+            PosixFileAttributeView view = Files.getFileAttributeView(target, PosixFileAttributeView.class);
+            if (view == null) {
+                return; // non-POSIX filesystem — nothing this JVM can read/set here
+            }
+            Files.setPosixFilePermissions(tmp, Files.getPosixFilePermissions(target));
+        } catch (IOException e) {
+            log.debug("cannot preserve permissions of '{}' onto its replacement", target, e);
+        }
+    }
+
     /** {@code s}, or {@code null} when {@code s} is null/blank — the charter-presence test used above. */
     private static String nonBlank(String s) {
         return (s == null || s.isBlank()) ? null : s;

fleetd/src/main/java/dev/ltms/fleet/member/CompositePeerLauncher.java

@@ -10,6 +10,7 @@ import dev.ltms.fleet.peer.PeerHandle;
 import dev.ltms.fleet.peer.PeerLauncher;
 import dev.ltms.fleet.peer.PeerUnreachableException;
 import dev.ltms.fleet.peer.SpawnRequest;
+import dev.ltms.fleet.placement.BackendOutagePolicy;
 import dev.ltms.fleet.placement.BackendQuarantine;
 import dev.ltms.fleet.placement.PlacementCandidate;
 import dev.ltms.fleet.placement.PlacementContext;
@@ -92,6 +93,18 @@ public final class CompositePeerLauncher implements PeerLauncher {
     /** CB-578 stage B: credential cooldown, checked before an explicit spawn and filtered into placement. */
     private final BackendQuarantine quarantine;
 
+    /**
+     * fleetd #201 Unit 5: credential cool-off after repeated backend errors, checked before an
+     * explicit spawn and filtered into placement — a SEPARATE, shorter-lived source from
+     * {@link #quarantine}. A never-{@code record}-called instance is naturally inert (its
+     * {@code remainingCoolOffSeconds} always returns empty), so back-compat constructors that predate
+     * this feature share one fixed instance rather than needing a {@code none()} sentinel.
+     */
+    private final BackendOutagePolicy outagePolicy;
+
+    /** The shared inert instance back-compat constructors wire in — never {@code record}-called. */
+    private static final BackendOutagePolicy NO_OUTAGE_POLICY = new BackendOutagePolicy(System::nanoTime);
+
     /**
      * CB-557: the role pools an unqualified spawn draws its candidates from. A supplier that yields
      * {@code null}, and an empty pool for a role, both fall back to every configured profile — the
@@ -130,7 +143,8 @@ public final class CompositePeerLauncher implements PeerLauncher {
                                  Map<String, FleetConfig.Profile> profileConfigs,
                                  PlacementPolicy placementPolicy,
                                  Function<String, Integer> liveCount) {
-        this(delegates, defaultProfile, profileConfigs, placementPolicy, liveCount, null, BackendQuarantine.none());
+        this(delegates, defaultProfile, profileConfigs, placementPolicy, liveCount, null,
+                BackendQuarantine.none(), NO_OUTAGE_POLICY);
     }
 
     /**
@@ -148,13 +162,14 @@ public final class CompositePeerLauncher implements PeerLauncher {
                                  PlacementPolicy placementPolicy,
                                  Function<String, Integer> liveCount,
                                  FleetConfig.Fleet fleet) {
-        this(delegates, defaultProfile, profileConfigs, placementPolicy, liveCount, fleet, BackendQuarantine.none());
+        this(delegates, defaultProfile, profileConfigs, placementPolicy, liveCount, fleet,
+                BackendQuarantine.none(), NO_OUTAGE_POLICY);
     }
 
     /**
-     * Production constructor with role pools and quarantine (CB-578 stage B). The full-featured
-     * non-reloading form; {@link #CompositePeerLauncher(List, String, Supplier, Function, BackendQuarantine)}
-     * is what {@code Fleetd.main} actually wires up.
+     * Production constructor with role pools and quarantine (CB-578 stage B), no cool-off (fleetd
+     * #201 Unit 5). Kept for callers that predate the cool-off feature; use the 8-arg overload below
+     * to wire a real {@link BackendOutagePolicy}.
      *
      * @param quarantine required — pass {@link BackendQuarantine#none()} for a caller that does not
      *                   want the feature, never a defaulting overload (CB-578 stage B's own rule).
@@ -166,17 +181,42 @@ public final class CompositePeerLauncher implements PeerLauncher {
                                  Function<String, Integer> liveCount,
                                  FleetConfig.Fleet fleet,
                                  BackendQuarantine quarantine) {
+        this(delegates, defaultProfile, profileConfigs, placementPolicy, liveCount, fleet, quarantine,
+                NO_OUTAGE_POLICY);
+    }
+
+    /**
+     * Production constructor with role pools, quarantine (CB-578 stage B), and cool-off (fleetd
+     * #201 Unit 5). The full-featured non-reloading form;
+     * {@link #CompositePeerLauncher(List, String, Supplier, Function, BackendQuarantine, BackendOutagePolicy)}
+     * is what {@code Fleetd.main} actually wires up.
+     *
+     * @param quarantine   required — pass {@link BackendQuarantine#none()} for a caller that does not
+     *                     want the feature, never a defaulting overload (CB-578 stage B's own rule).
+     * @param outagePolicy required — pass a fresh, never-{@code record}-called {@link
+     *                     BackendOutagePolicy} for a caller that does not want the feature; same
+     *                     "explicit opt-out, never a silent default" rule as {@code quarantine}.
+     */
+    public CompositePeerLauncher(List<HerdrPeerLauncher> delegates,
+                                 String defaultProfile,
+                                 Map<String, FleetConfig.Profile> profileConfigs,
+                                 PlacementPolicy placementPolicy,
+                                 Function<String, Integer> liveCount,
+                                 FleetConfig.Fleet fleet,
+                                 BackendQuarantine quarantine,
+                                 BackendOutagePolicy outagePolicy) {
         // LinkedHashMap, not Map.copyOf: candidates() promises definition order and the weighted
         // policy breaks exact-weight ties on it, so a salted iteration order would make placement
         // differ from one JVM run to the next.
         this(delegates, defaultProfile,
                 constant(Collections.unmodifiableMap(new LinkedHashMap<>(profileConfigs))),
-                constant(placementPolicy), liveCount, constant(fleet), quarantine);
+                constant(placementPolicy), liveCount, constant(fleet), quarantine, outagePolicy);
     }
 
     /**
      * Production constructor that re-reads its placement inputs per spawn (CB-559), so a config
-     * reload retargets the next member without a restart.
+     * reload retargets the next member without a restart. No cool-off (fleetd #201 Unit 5); use the
+     * 6-arg overload below to wire a real {@link BackendOutagePolicy}.
      *
      * @param config     the live configuration — read at every spawn, never captured
      * @param quarantine required — CB-578 stage B; pass {@link BackendQuarantine#none()} to opt out
@@ -186,12 +226,31 @@ public final class CompositePeerLauncher implements PeerLauncher {
                                  Supplier<FleetConfig> config,
                                  Function<String, Integer> liveCount,
                                  BackendQuarantine quarantine) {
+        this(delegates, defaultProfile, config, liveCount, quarantine, NO_OUTAGE_POLICY);
+    }
+
+    /**
+     * Production constructor that re-reads its placement inputs per spawn (CB-559), with cool-off
+     * (fleetd #201 Unit 5). This is what {@code Fleetd.main} actually wires up.
+     *
+     * @param config       the live configuration — read at every spawn, never captured
+     * @param quarantine   required — CB-578 stage B; pass {@link BackendQuarantine#none()} to opt out
+     * @param outagePolicy required — fleetd #201 Unit 5; pass a fresh, never-{@code record}-called
+     *                     {@link BackendOutagePolicy} to opt out
+     */
+    public CompositePeerLauncher(List<HerdrPeerLauncher> delegates,
+                                 String defaultProfile,
+                                 Supplier<FleetConfig> config,
+                                 Function<String, Integer> liveCount,
+                                 BackendQuarantine quarantine,
+                                 BackendOutagePolicy outagePolicy) {
         this(delegates, defaultProfile,
                 () -> config.get().profiles(),
                 () -> PlacementPolicies.fromName(config.get().placement()),
                 liveCount,
                 () -> config.get().fleet(),
-                quarantine);
+                quarantine,
+                outagePolicy);
     }
 
     /** The all-suppliers form every other constructor funnels into. */
@@ -201,9 +260,11 @@ public final class CompositePeerLauncher implements PeerLauncher {
                                   Supplier<PlacementPolicy> placementPolicy,
                                   Function<String, Integer> liveCount,
                                   Supplier<FleetConfig.Fleet> fleet,
-                                  BackendQuarantine quarantine) {
+                                  BackendQuarantine quarantine,
+                                  BackendOutagePolicy outagePolicy) {
         this.fleet = fleet;
         this.quarantine = Objects.requireNonNull(quarantine, "quarantine");
+        this.outagePolicy = Objects.requireNonNull(outagePolicy, "outagePolicy");
         if (delegates.isEmpty()) {
             throw new IllegalArgumentException("at least one peer adapter must be configured");
         }
@@ -266,7 +327,10 @@ public final class CompositePeerLauncher implements PeerLauncher {
             // the charter makes explicit-profile spawns the normal path — so skipping the check
             // here would leave the cap dead config in real operation.
             HerdrPeerLauncher d = route(requestedProfile);
+            // Checked in this order so exhaustion quarantine wins when both are active: quarantine
+            // throws first and short-circuits before the cool-off check ever runs (fleetd #201 Unit 5).
             enforceNotQuarantined(requestedProfile);
+            enforceNotCoolingOff(requestedProfile);
             enforceMaxLoad(requestedProfile);
             PeerHandle handle = d.spawn(req);
             spawnedBy.put(handle.id(), d);
@@ -283,7 +347,10 @@ public final class CompositePeerLauncher implements PeerLauncher {
         // CB-578 stage B: computed once up front — a quarantine's expiry cannot pass within one spawn
         // call, so re-deriving it per retry would only cost work, never change the answer.
         Set<String> quarantined = quarantinedProfiles(candidates);
-        PlacementContext ctx = new PlacementContext(roleDefault, candidates, liveCount, unreachable, quarantined);
+        // fleetd #201 Unit 5: a distinct set from quarantined — see PlacementContext.coolingOff.
+        Set<String> coolingOff = coolingOffProfiles(candidates);
+        PlacementContext ctx = new PlacementContext(roleDefault, candidates, liveCount, unreachable,
+                quarantined, coolingOff);
 
         int maxAttempts = candidates.isEmpty() ? 1 : candidates.size();
         for (int attempt = 0; attempt < maxAttempts; attempt++) {
@@ -313,7 +380,8 @@ public final class CompositePeerLauncher implements PeerLauncher {
                         chosen.profile(), e.getMessage());
                 unreachable.add(chosen.profile());
                 // Update the context for the next selection so the policy excludes this profile.
-                ctx = new PlacementContext(roleDefault, candidates, liveCount, unreachable, quarantined);
+                ctx = new PlacementContext(roleDefault, candidates, liveCount, unreachable,
+                        quarantined, coolingOff);
             }
         }
 
@@ -379,6 +447,31 @@ public final class CompositePeerLauncher implements PeerLauncher {
                 .collect(Collectors.toSet());
     }
 
+    /**
+     * Refuse an explicit-profile spawn whose credential is cooling off after repeated backend errors
+     * (fleetd #201 Unit 5 — {@link BackendOutagePolicy}): a SEPARATE, shorter-lived source from
+     * {@link #enforceNotQuarantined}'s exhaustion quarantine. Checked after quarantine so exhaustion
+     * wins when both are active — see the call site in {@link #spawn}.
+     *
+     * @throws PlacementException naming the profile, its credential, and the remaining cool-off
+     */
+    private void enforceNotCoolingOff(String profile) {
+        String credentialId = credentialIdFor(profile);
+        outagePolicy.remainingCoolOffSeconds(credentialId).ifPresent(remaining -> {
+            throw new PlacementException("worker profile '" + profile + "' credential '" + credentialId
+                    + "' is cooling off after repeated backend errors; ~" + remaining
+                    + "s remaining — refusing spawn");
+        });
+    }
+
+    /** The subset of {@code candidates} whose credential is currently cooling off (fleetd #201 Unit 5). */
+    private Set<String> coolingOffProfiles(List<PlacementCandidate> candidates) {
+        return candidates.stream()
+                .map(PlacementCandidate::profile)
+                .filter(p -> outagePolicy.remainingCoolOffSeconds(credentialIdFor(p)).isPresent())
+                .collect(Collectors.toSet());
+    }
+
     private void enforceMaxLoad(String profile) {
         // Absent config, or a config whose maxLoad normalized to null (ABSENT ⇒ unlimited at load),
         // means no cap — never cap what wasn't configured. Note "non-positive ⇒ unlimited" was true

fleetd/src/main/java/dev/ltms/fleet/member/HerdrPeerLauncher.java

@@ -358,6 +358,42 @@ public abstract class HerdrPeerLauncher implements PeerLauncher {
         return Files.isRegularFile(candidate) ? candidate : null;
     }
 
+    /**
+     * Whether {@code cwd} is a fleetd-provisioned git worktree — signalled the same way
+     * {@code ClaudeCodeLauncher#writeIdeOverlay} already gates on: a {@code .git} that is a
+     * <strong>regular file</strong> holding a {@code gitdir:} pointer, as opposed to a real
+     * checkout's {@code .git} <strong>directory</strong>. {@code null}/blank never qualifies.
+     *
+     * <p>Shared by every write (and, since fleetd #249, every identity read) that must land only
+     * in a worktree fleetd itself created for a member — never in a real checkout, an arbitrary
+     * configured directory, or (see the incident below) the daemon's own fallback cwd. Package-
+     * private (not {@code protected}) on purpose: {@link ClaudeCodeLauncher} and
+     * {@link OpenCodeLauncher} both call it, and same-package visibility is enough — no subclass
+     * outside this package needs it.
+     *
+     * <p><b>fleetd #149 incident.</b> {@code ClaudeCodeLauncher#seedTrustDialog} originally ran
+     * unconditionally on any non-blank {@code cwd}. Most of that launcher's OWN tests spawn a
+     * profile with no {@code cwd} configured, so the base class's {@code resolveCwd} falls
+     * through to the real {@code user.dir} — and with no {@code configDir} either (also the
+     * common case in that file's fixtures), the seed's target falls through the same way to the
+     * real {@code ~/.claude.json}. Running this repo's own test suite corrupted the operator's
+     * actual config file (it shrank from ~72 KB to a single seeded entry) the first time a
+     * mutation happened to make the write non-additive. Gating both cwd-targeted writes on "this
+     * is a worktree fleetd provisioned" — exactly the population fleetd #149 describes
+     * ({@code worktree: true} always lands in a brand-new directory) — makes that class of write
+     * impossible against a real checkout or an untouched fallback cwd, in production or in tests.
+     *
+     * <p><b>fleetd #249.</b> The same reasoning extends to a READ: {@code
+     * OpenCodeSessionDiscovery#sessionIdForDirectory} keys on {@code directory}, a heuristic that
+     * is only reliable when the directory is unique to this member — i.e., exactly the population
+     * this gate identifies. {@link OpenCodeLauncher} uses it to withhold {@code agentSessionId()}
+     * (report absence rather than a guess) and to refuse a {@code resumeSessionId} spawn that
+     * cannot be resolved reliably going forward.
+     */
+    static boolean isProvisionedWorktree(String cwd) {
+        return cwd != null && !cwd.isBlank() && Files.isRegularFile(Path.of(cwd, ".git"));
+    }
+
     // --- profile surface -----------------------------------------------------------------------
 
     /** The configured peer profile names (what {@code spawn(profile)} accepts). */

fleetd/src/main/java/dev/ltms/fleet/member/MemberCredentialPolicyView.java

@@ -0,0 +1,75 @@
+package dev.ltms.fleet.member;
+
+import dev.ltms.fleet.config.FleetConfig;
+
+import java.util.List;
+
+/**
+ * fleetd #111 (CB-608): a single, testable read of the {@code memberCredentials:} policy — names
+ * and counts only, never a value. The daemon never holds a credential's <em>value</em> in the
+ * first place (only the names configured under {@code known:}/{@code allow:}), so there is
+ * nothing here to redact by construction; the point of this class is that it is the ONE place
+ * that turns a policy into names-and-counts, so nothing else hand-counts a second time.
+ *
+ * <p>Before this class, {@link dev.ltms.fleet.Fleetd#reportMemberCredentialsGap} computed these
+ * same counts inline for the startup log line, and {@code scripts/probe-member-credentials.sh}
+ * carried its own hardcoded {@code NAMES} array that the live policy could grow past silently
+ * (#111) — the exact "hand-maintained second copy drifts" shape #114 fixed for the tool
+ * catalogue. Both now read this class: the startup log via {@link
+ * dev.ltms.fleet.Fleetd#reportMemberCredentialsGap}, and a live daemon via the {@code
+ * GET /member-credentials} REST endpoint ({@link dev.ltms.fleet.rest.FleetApp}), which the probe
+ * script fetches instead of carrying its own list.
+ *
+ * @param present policy configured with at least one {@code known} name. {@code false} for an
+ *                absent or empty {@code memberCredentials:} block — represented honestly as "no
+ *                policy", never as "nothing blocked" (an empty {@link #blocked} could otherwise be
+ *                misread as a clean bill of health).
+ * @param policy  the normalized policy mode ({@link FleetConfig.MemberCredentials#policy()}), or
+ *                {@code null} when {@link #present} is {@code false}.
+ * @param known   every name the policy declares, in configured order. Names only, never a value.
+ * @param allowed the subset of {@link #known} explicitly let through. Names only.
+ * @param blocked {@link #known} minus {@link #allowed} — the names an actual spawn shadows. Names
+ *                only.
+ */
+public record MemberCredentialPolicyView(boolean present, String policy, List<String> known,
+                                          List<String> allowed, List<String> blocked) {
+
+    private static final MemberCredentialPolicyView ABSENT =
+            new MemberCredentialPolicyView(false, null, List.of(), List.of(), List.of());
+
+    public MemberCredentialPolicyView {
+        known = known == null ? List.of() : List.copyOf(known);
+        allowed = allowed == null ? List.of() : List.copyOf(allowed);
+        blocked = blocked == null ? List.of() : List.copyOf(blocked);
+    }
+
+    /** The honest "no policy configured" view. */
+    public static MemberCredentialPolicyView absent() {
+        return ABSENT;
+    }
+
+    /**
+     * Build the view straight from the live config. {@code creds} may be {@code null} (no {@code
+     * memberCredentials:} block at all) — treated the same as a present-but-empty block, exactly
+     * like {@link dev.ltms.fleet.Fleetd#reportMemberCredentialsGap} already did.
+     */
+    public static MemberCredentialPolicyView of(FleetConfig.MemberCredentials creds) {
+        if (creds == null || creds.known().isEmpty()) {
+            return ABSENT;
+        }
+        return new MemberCredentialPolicyView(true, creds.policy(), creds.known(), creds.allow(),
+                List.copyOf(creds.blockedSet()));
+    }
+
+    public int knownCount() {
+        return known.size();
+    }
+
+    public int allowedCount() {
+        return allowed.size();
+    }
+
+    public int blockedCount() {
+        return blocked.size();
+    }
+}

fleetd/src/main/java/dev/ltms/fleet/member/OpenCodeLauncher.java

@@ -672,13 +672,29 @@ public final class OpenCodeLauncher extends HerdrPeerLauncher {
     /** Add lazy on-disk session discovery to the base handle. */
     @Override
     public PeerHandle spawn(SpawnRequest req) {
+        String cwd = effectiveCwd(req);
+        // fleetd #249: refuse rather than silently resume into unverifiable territory. opencode's
+        // `-s <id>` flag itself resumes precisely — the resolved id is what fails, not the resume —
+        // but resolvedSessionId() below can never confirm (or later re-report) this handle's own
+        // identity without a fleetd-provisioned worktree (isProvisionedWorktree(cwd)), because the
+        // directory is shared and sessionIdForDirectory's "most recently updated row" heuristic can
+        // pick a sibling's session. Refusing here, before anything spawns, beats letting the member
+        // start and only then discovering fleetd can never again verify who it actually is.
+        if (req.resumeSessionId() != null && !req.resumeSessionId().isBlank()
+                && !isProvisionedWorktree(cwd)) {
+            throw new IllegalArgumentException("resumeSessionId requires a fleetd-provisioned "
+                    + "worktree for an opencode profile — without one, this member's cwd is shared "
+                    + "with other sessions, so fleetd can never reliably confirm (now or later) which "
+                    + "conversation it is actually running (fleetd #249). Pass fleet_spawn{worktree:"
+                    + "<ticket-slug>} to resume this member.");
+        }
         PeerHandle inner = super.spawn(req);
         // fleetd #175: the same profile config buildLaunch resolved for this spawn (requireProfile
         // is deterministic on req.profileName(), so re-resolving here costs a map lookup, not a
         // second decision) — SessionAwareHandle needs cfg.model() to know what THIS session should
         // be running.
         FleetConfig.Profile cfg = requireProfile(req.profileName());
-        return new SessionAwareHandle(inner, discovery, effectiveCwd(req), cfg,
+        return new SessionAwareHandle(inner, discovery, cwd, cfg,
                 this::memberHerdrSocketConfigured, discoveryUnavailableWarned, exhaustionSink);
     }
 
@@ -718,6 +734,17 @@ public final class OpenCodeLauncher extends HerdrPeerLauncher {
          * the directory right now."
          */
         private final AtomicReference<String> resolvedSessionId = new AtomicReference<>();
+        /**
+         * fleetd #249: whether {@link #cwd} is a fleetd-provisioned git worktree
+         * ({@link HerdrPeerLauncher#isProvisionedWorktree}), computed once at spawn time since
+         * {@code cwd} never changes for this handle. When {@code false} the directory is shared
+         * with other sessions (the default no-worktree spawn inherits the lead's own cwd), so
+         * {@link OpenCodeSessionDiscovery#sessionIdForDirectory}'s "most recently updated row for
+         * this directory" heuristic can and does pick another session's row — see that class's
+         * javadoc. {@link #agentSessionId()} refuses to guess in that case: it reports absent
+         * rather than a possibly-foreign id.
+         */
+        private final boolean worktreeProvisioned;
 
         SessionAwareHandle(PeerHandle delegate, OpenCodeSessionDiscovery discovery, String cwd,
                             FleetConfig.Profile cfg,
@@ -731,6 +758,7 @@ public final class OpenCodeLauncher extends HerdrPeerLauncher {
             this.discoveryUnavailable = discoveryUnavailable;
             this.discoveryUnavailableWarned = discoveryUnavailableWarned;
             this.exhaustionSink = exhaustionSink;
+            this.worktreeProvisioned = isProvisionedWorktree(cwd);
         }
 
         @Override
@@ -760,6 +788,9 @@ public final class OpenCodeLauncher extends HerdrPeerLauncher {
             // built from (see OpenCodeLauncher#defaultDiscoveryRoot's javadoc for the full
             // reasoning). Scanning fleetd's own $HOME under that config would only ever find "no
             // row" and read as "resume unsupported" — declare it unavailable instead, once, loudly.
+            // Checked before the fleetd #249 worktree gate below: this OS-user mismatch makes
+            // discovery unusable regardless of whether cwd happens to be a provisioned worktree, so
+            // it earns the one-time WARN either way.
             if (discoveryUnavailable.getAsBoolean()) {
                 if (discoveryUnavailableWarned.compareAndSet(false, true)) {
                     log.warn("opencode session discovery unavailable: memberHerdrSocket is "
@@ -771,6 +802,16 @@ public final class OpenCodeLauncher extends HerdrPeerLauncher {
                 }
                 return null;
             }
+            // fleetd #249: cwd is shared with other sessions unless fleetd itself provisioned this
+            // worktree, and sessionIdForDirectory's directory-keyed heuristic cannot tell this
+            // member's row apart from a sibling's in that case (measured: a three-day-old row from
+            // a different profile). Refuse to guess — absent is the honest answer, and it is what
+            // this codebase already returns elsewhere for absent evidence (fleetd #175's UNKNOWN).
+            // No WARN here: unlike discoveryUnavailable above, this is the ordinary, expected shape
+            // of the large majority of spawns (no worktree requested), not a configuration gap.
+            if (!worktreeProvisioned) {
+                return null;
+            }
             // fleetd #234: once resolved, stay resolved. Re-deriving from `directory` on every call
             // would let this handle's identity drift to a sibling session that later shares the
             // same cwd and writes a newer row — see resolvedSessionId's javadoc.

fleetd/src/main/java/dev/ltms/fleet/placement/FixedPlacementPolicy.java

@@ -1,55 +1,68 @@
 package dev.ltms.fleet.placement;
 
+import java.util.ArrayList;
+import java.util.List;
+
 /**
  * Backward-compatible placement: an unqualified spawn always resolves to the configured default
  * profile, exactly as {@code CompositePeerLauncher} did before CB-518. This ignores caps and
  * reachability so that a pre-existing config behaves identically after upgrade.
  *
- * <p>Two exceptions walk past the default instead of returning it unconditionally:
+ * <p>Three exceptions walk past the default instead of returning it unconditionally:
  * <ul>
  *   <li>Quarantine (CB-578 stage B): a quarantined default is a credential that just refused on
  *       a usage limit, not a transient capacity or reachability concern.
+ *   <li>Cooling off (fleetd #201 Unit 5): a credential cooling off after repeated backend errors
+ *       ({@code BackendOutagePolicy}) — a separate, shorter-lived source from quarantine. When a
+ *       profile is both quarantined and cooling off, only the quarantine reason is reported
+ *       (exhaustion takes priority), matching {@code CompositePeerLauncher}'s explicit-spawn order.
  *   <li>Weight 0 (CB-554): {@code fixed} is still automatic selection, so a profile the operator
  *       marked "never auto-select me" ({@code weight <= 0}) must be skipped here exactly as
  *       {@code weighted}/{@code round-robin} skip it — an explicit {@code fleet_spawn} naming
  *       the profile is unaffected, only this automatic fallback walk.
  * </ul>
- * A fleet where nothing is ever quarantined or weight-0 never exercises either path, so today's
- * behaviour is unchanged.
+ * A fleet where nothing is ever quarantined, cooling off, or weight-0 never exercises any of these
+ * paths, so today's behaviour is unchanged.
  */
 final class FixedPlacementPolicy implements PlacementPolicy {
 
     @Override
     public PlacementCandidate select(PlacementContext ctx) {
         String d = ctx.defaultProfile();
-        if (d != null && !d.isBlank() && !ctx.quarantined().contains(d) && !weightExcluded(ctx, d)) {
+        if (d != null && !d.isBlank() && !ctx.quarantined().contains(d) && !ctx.coolingOff().contains(d)
+                && !weightExcluded(ctx, d)) {
             return new PlacementCandidate(d, null, 1.0f, null);
         }
         for (PlacementCandidate c : ctx.candidates()) {
-            if (!ctx.quarantined().contains(c.profile()) && !c.excluded()) {
+            if (!ctx.quarantined().contains(c.profile()) && !ctx.coolingOff().contains(c.profile())
+                    && !c.excluded()) {
                 return new PlacementCandidate(c.profile(), null, c.weight(), c.maxLoad());
             }
         }
         if (d != null && !d.isBlank()) {
             boolean dQuarantined = ctx.quarantined().contains(d);
+            // Exhaustion quarantine takes priority: reported only when quarantine is absent, so the
+            // message never claims "cooling off" for a profile that is really backend-exhausted.
+            boolean dCoolingOff = !dQuarantined && ctx.coolingOff().contains(d);
             boolean dWeightExcluded = weightExcluded(ctx, d);
-            if (dQuarantined && dWeightExcluded) {
-                throw new PlacementException("worker profile '" + d + "' is quarantined (backend "
-                        + "exhausted) and has weight 0 (excluded from automatic selection), and no "
-                        + "available candidate remains");
-            }
-            if (dWeightExcluded) {
-                throw new PlacementException("worker profile '" + d + "' has weight 0 (excluded "
-                        + "from automatic selection) and no available candidate remains");
-            }
-            if (dQuarantined) {
-                throw new PlacementException("worker profile '" + d + "' is quarantined (backend "
-                        + "exhausted) and no un-quarantined candidate is available");
+            if (dQuarantined || dCoolingOff || dWeightExcluded) {
+                List<String> reasons = new ArrayList<>();
+                if (dQuarantined) {
+                    reasons.add("is quarantined (backend exhausted)");
+                }
+                if (dCoolingOff) {
+                    reasons.add("is cooling off after repeated backend errors");
+                }
+                if (dWeightExcluded) {
+                    reasons.add("has weight 0 (excluded from automatic selection)");
+                }
+                throw new PlacementException("worker profile '" + d + "' "
+                        + String.join(" and ", reasons) + ", and no available candidate remains");
             }
         }
         if (!ctx.candidates().isEmpty()) {
-            throw new PlacementException(
-                    "all worker profiles are excluded from automatic selection (quarantined or weight-0)");
+            throw new PlacementException("all worker profiles are excluded from automatic "
+                    + "selection (quarantined, cooling off, or weight-0)");
         }
         throw new PlacementException("no worker profiles configured");
     }

fleetd/src/main/java/dev/ltms/fleet/placement/PlacementContext.java

@@ -14,10 +14,19 @@ import java.util.function.Function;
  * @param quarantined    profiles whose credential is currently quarantined (CB-578 stage B) — a
  *                       {@code BACKEND_EXHAUSTED} classification put it, or a profile it shares a
  *                       credential with, on cooldown. Filtered the same way as {@code unreachable}.
+ * @param coolingOff     profiles whose credential is currently cooling off after repeated backend
+ *                       errors (fleetd #201 Unit 5 — {@code BackendOutagePolicy}), a SEPARATE,
+ *                       shorter-lived source from {@code quarantined}: a credential outage cools off
+ *                       even when no member was ever exhausted. Deliberately its own set rather than
+ *                       merged into {@code quarantined} — {@link PlacementPolicyUtil} needs to tell
+ *                       the two apart so its refusal message says "cooling off", not "exhausted",
+ *                       when only this one is active. A profile can be in both sets at once; when it
+ *                       is, exhaustion quarantine is reported (it takes priority).
  */
 public record PlacementContext(String defaultProfile,
                                List<PlacementCandidate> candidates,
                                Function<String, Integer> liveCount,
                                Set<String> unreachable,
-                               Set<String> quarantined) {
+                               Set<String> quarantined,
+                               Set<String> coolingOff) {
 }

fleetd/src/main/java/dev/ltms/fleet/placement/PlacementPolicyUtil.java

@@ -14,14 +14,16 @@ final class PlacementPolicyUtil {
     /**
      * Candidates that are not weight-excluded (CB-554: explicit {@code weight <= 0}, checked
      * first because it is a static config choice rather than transient state), not
-     * known-unreachable, not quarantined (CB-578 stage B), and have not reached their maxLoad.
-     * A {@code null} maxLoad means unlimited.
+     * known-unreachable, not quarantined (CB-578 stage B), not cooling off after repeated backend
+     * errors (fleetd #201 Unit 5 — a separate, shorter-lived source from quarantine), and have not
+     * reached their maxLoad. A {@code null} maxLoad means unlimited.
      */
     static List<PlacementCandidate> available(PlacementContext ctx) {
         List<PlacementCandidate> out = new ArrayList<>();
         for (PlacementCandidate c : ctx.candidates()) {
             if (c.excluded() || ctx.unreachable().contains(c.profile())
-                    || ctx.quarantined().contains(c.profile())) {
+                    || ctx.quarantined().contains(c.profile())
+                    || ctx.coolingOff().contains(c.profile())) {
                 continue;
             }
             Integer cap = c.maxLoad();
@@ -38,21 +40,27 @@ final class PlacementPolicyUtil {
 
     /**
      * Build a clear exception describing why every candidate was dropped: all weight-0, all
-     * quarantined, all at capacity, all unreachable, or a mix. Each candidate is counted into
-     * exactly one bucket (weight-excluded takes priority) so a candidate excluded for more than
-     * one reason is never double-counted.
+     * quarantined, all cooling off, all at capacity, all unreachable, or a mix. Each candidate is
+     * counted into exactly one bucket (weight-excluded first, then quarantined, then cooling off)
+     * so a candidate excluded for more than one reason is never double-counted — a candidate that is
+     * both quarantined (CB-578 stage B, backend exhausted) and cooling off (fleetd #201 Unit 5,
+     * repeated backend errors) counts only as quarantined, matching {@code CompositePeerLauncher}'s
+     * explicit-spawn ordering: exhaustion quarantine takes priority when both are active.
      */
     static PlacementException emptyException(PlacementContext ctx) {
         int weightExcluded = 0;
         int atCap = 0;
         int unreachable = 0;
         int quarantined = 0;
+        int coolingOff = 0;
         for (PlacementCandidate c : ctx.candidates()) {
             Integer cap = c.maxLoad();
             if (c.excluded()) {
                 weightExcluded++;
             } else if (ctx.quarantined().contains(c.profile())) {
                 quarantined++;
+            } else if (ctx.coolingOff().contains(c.profile())) {
+                coolingOff++;
             } else if (ctx.unreachable().contains(c.profile())) {
                 unreachable++;
             } else if (cap != null && ctx.liveCount().apply(c.profile()) >= cap) {
@@ -71,6 +79,10 @@ final class PlacementPolicyUtil {
         if (quarantined == total) {
             return new PlacementException("all worker profiles are quarantined (backend exhausted)");
         }
+        if (coolingOff == total) {
+            return new PlacementException(
+                    "all worker profiles are cooling off after repeated backend errors");
+        }
         if (atCap == total) {
             return new PlacementException("all worker profiles are at maxLoad");
         }
@@ -79,7 +91,9 @@ final class PlacementPolicyUtil {
         }
         return new PlacementException("no worker profile available: " + atCap + " at maxLoad, "
                 + unreachable + " unreachable, " + quarantined + " quarantined, "
+                + coolingOff + " cooling off, "
                 + weightExcluded + " weight-0, "
-                + (total - atCap - unreachable - quarantined - weightExcluded) + " remaining");
+                + (total - atCap - unreachable - quarantined - coolingOff - weightExcluded)
+                + " remaining");
     }
 }

fleetd/src/main/java/dev/ltms/fleet/rest/FleetApp.java

@@ -13,6 +13,7 @@ import dev.ltms.fleet.herdr.Agent;
 import dev.ltms.fleet.herdr.HerdrClient;
 import dev.ltms.fleet.herdr.HerdrException;
 import dev.ltms.fleet.inject.MemberPresence;
+import dev.ltms.fleet.member.MemberCredentialPolicyView;
 import dev.ltms.fleet.peer.PeerUnreachableException;
 import dev.ltms.fleet.placement.PlacementException;
 import dev.ltms.fleet.msg.MessageService;
@@ -32,6 +33,7 @@ import java.util.List;
 import java.util.Map;
 import java.util.function.Function;
 import java.util.function.Predicate;
+import java.util.function.Supplier;
 import java.util.stream.Collectors;
 
 /**
@@ -64,6 +66,10 @@ public final class FleetApp {
     private final HttpServlet mcpServlet; // MCP Streamable-HTTP endpoint, mounted at /mcp (nullable)
     private final CallerResolver auth;    // CB-501: null → authz not enforced (legacy behaviour)
     private final Metrics metrics;        // CB-502: null → /metrics not exposed
+    // fleetd #111: re-read per request, same hot-reload shape as every other live config read —
+    // absent() (the honest "no policy configured" view) for every constructor that does not wire
+    // a real one, so existing legacy call sites keep building without knowing this field exists.
+    private final Supplier<MemberCredentialPolicyView> memberCredentials;
     private final ObjectMapper mapper = new ObjectMapper();
 
     /**
@@ -111,6 +117,19 @@ public final class FleetApp {
                        MessageService messages, MemberPresence presence,
                        HttpServlet mcpServlet, CallerResolver auth, Metrics metrics,
                        Predicate<String> deliverable) {
+        this(herdr, memberHerdr, workers, sessions, messages, presence, mcpServlet, auth, metrics,
+                deliverable, MemberCredentialPolicyView::absent);
+    }
+
+    /**
+     * @param memberCredentials live {@code memberCredentials:} policy view (fleetd #111), re-read
+     *                          per request for {@code GET /member-credentials}; production wiring
+     *                          passes the same hot-reload shape as every other live config read
+     */
+    public FleetApp(HerdrClient herdr, HerdrClient memberHerdr, PeerLauncher workers, SessionManager sessions,
+                       MessageService messages, MemberPresence presence,
+                       HttpServlet mcpServlet, CallerResolver auth, Metrics metrics,
+                       Predicate<String> deliverable, Supplier<MemberCredentialPolicyView> memberCredentials) {
         this.herdr = herdr;
         this.memberHerdr = memberHerdr != null ? memberHerdr : herdr;
         this.workers = workers;
@@ -120,6 +139,7 @@ public final class FleetApp {
         this.mcpServlet = mcpServlet;
         this.auth = auth;
         this.metrics = metrics;
+        this.memberCredentials = memberCredentials != null ? memberCredentials : MemberCredentialPolicyView::absent;
     }
 
     /** Wire routes onto a fresh, unstarted Javalin instance. Caller starts it. */
@@ -148,6 +168,7 @@ public final class FleetApp {
         app.get("/agents", this::agents);
         app.get("/members", this::listMembers);                  // CB-304: registry roster + live herdr status
         app.get("/profiles", this::profiles);                    // configured backend profiles
+        app.get("/member-credentials", this::memberCredentials); // fleetd #111: policy names + counts, never a value
         app.post("/members", this::spawnMember);                 // optional ?role=&profile= or {"role":…,"profile":…}
         app.delete("/members/{paneId}", this::stopMember);
         app.post("/sessions/{id}/message", this::sendMessage);   // fleet_send (primary; blocking, wait:false, or answer via turnId)
@@ -346,6 +367,29 @@ public final class FleetApp {
                 "default", workers.defaultProfile() == null ? "" : workers.defaultProfile()));
     }
 
+    /**
+     * fleetd #111 (CB-608): the live {@code memberCredentials:} policy as names and counts —
+     * NEVER a value. The daemon does not hold a credential's value in the first place (only the
+     * name it is configured under), so there is nothing to redact here beyond what {@link
+     * MemberCredentialPolicyView} already omits by construction. This is the source
+     * {@code scripts/probe-member-credentials.sh} reads instead of carrying its own hardcoded
+     * name list, which is exactly what let the list drift silently behind the real policy.
+     */
+    private void memberCredentials(Context ctx) {
+        if (!allow(ctx, Authz.Action.READ, null)) {
+            return;
+        }
+        MemberCredentialPolicyView view = memberCredentials.get();
+        ctx.status(200).json(Map.of(
+                "present", view.present(),
+                "policy", view.policy() == null ? "" : view.policy(),
+                "known", view.known(),
+                "allowed", view.allowed(),
+                "knownCount", view.knownCount(),
+                "allowedCount", view.allowedCount(),
+                "blockedCount", view.blockedCount()));
+    }
+
     /**
      * Spawn a guard-checked worker. An optional {@code profile} (query param or {@code {"profile":…}}
      * body) picks which configured profile; omitted → the default. 403 if the base_url would breach

fleetd/src/main/java/dev/ltms/fleet/session/GitWorktrees.java

@@ -426,19 +426,75 @@ public final class GitWorktrees implements Worktrees {
      * neutralized copy from ever showing up as a local modification the worker might commit. A config
      * the repo does not carry is skipped silently — no stub is invented for a file the repo does not
      * have, and one missing file must never fail provisioning.
+     *
+     * <p>fleetd #134. The neutralization above is correct and stays unconditional — the defect was
+     * that it was invisible on both sides. Neither the daemon's own log nor the worker sitting in the
+     * worktree could tell a stub from the repo's real file: a real worker read a 3-byte {@code {}}
+     * where the repo's {@code opencode.json} is 30+ lines, and reported — truthfully from what it
+     * could see, and wrongly — that a mount key did not exist. Two fixes, aimed at two different
+     * readers:
+     * <ul>
+     *   <li>the daemon operator reads {@link #log}, so the summary below names the denominator, what
+     *       was neutralized, and why anything was not — the same shape {@code overlayParity} reports
+     *       its own copy in;
+     *   <li>the worker reads its own worktree, not the daemon's log, so the same fact is recorded a
+     *       second time in worktree-scoped git config ({@code fleet.neutralizedConfig} /
+     *       {@code fleet.neutralizedConfigNote}, readable with {@code git config --worktree --get-all
+     *       fleet.neutralizedConfig}) rather than as a file in the working tree. A working-tree file
+     *       would show up in {@code git status} for the worker to trip on or commit; worktree-scoped
+     *       config lives in {@code .git/worktrees/<nonce>/config.worktree} and can never appear there.
+     *       This reuses the exact mechanism {@link #configureEnvironmentCredentialHelper} and
+     *       {@link #configureHttpsUrlRewriteForSshOrigin} already use for other worktree-local state.
+     * </ul>
      */
     private void isolateToolSurface(String worktreePath) {
         Path root = Path.of(worktreePath).toAbsolutePath().normalize();
+        List<String> neutralized = new ArrayList<>();
+        List<String> skipped = new ArrayList<>();
         for (WorktreeHostileConfig cfg : WORKTREE_HOSTILE_CONFIGS) {
-            neutralize(root, worktreePath, cfg);
+            if (neutralize(root, worktreePath, cfg)) {
+                neutralized.add(cfg.file());
+            } else {
+                skipped.add(cfg.file() + " absent");
+            }
         }
+        String detail = neutralized.isEmpty() ? String.join(", ", skipped)
+                : skipped.isEmpty() ? String.join(", ", neutralized)
+                : String.join(", ", neutralized) + " (" + String.join(", ", skipped) + ")";
+        log.info("tool-surface isolation: neutralized {} of {} configs: {} — the worktree copy is a "
+                        + "stub, not the repo's file; edit the real file in the primary checkout instead",
+                neutralized.size(), WORKTREE_HOSTILE_CONFIGS.size(), detail);
+        recordNeutralizedConfigForWorker(worktreePath, neutralized);
     }
 
-    private void neutralize(Path root, String worktreePath, WorktreeHostileConfig cfg) {
+    /**
+     * The worker-readable half of fleetd #134: record which files were neutralized where the worker
+     * itself can read it, without a working-tree file that would show up in {@code git status}.
+     * Worktree-scoped git config is per-worktree, lives under {@code .git/worktrees/<nonce>/} rather
+     * than the working tree, and this repo already relies on the same mechanism (and the same
+     * {@code extensions.worktreeConfig} enablement) for the credential helper and the SSH→HTTPS
+     * rewrite — see {@link #configureEnvironmentCredentialHelper}.
+     */
+    private void recordNeutralizedConfigForWorker(String worktreePath, List<String> neutralized) {
+        if (neutralized.isEmpty()) {
+            return;
+        }
+        exec("git", "-C", worktreePath, "config", "extensions.worktreeConfig", "true");
+        for (String file : neutralized) {
+            exec("git", "-C", worktreePath, "config", "--worktree", "--add", "fleet.neutralizedConfig", file);
+        }
+        exec("git", "-C", worktreePath, "config", "--worktree", "fleet.neutralizedConfigNote",
+                "the worktree copy of each fleet.neutralizedConfig path is a stub, not the repo's "
+                        + "committed file; edit the real file from the primary checkout instead");
+    }
+
+    /** @return true if {@code cfg} was neutralized (present, or created because {@link
+     *  WorktreeHostileConfig#createIfAbsent()}); false if the repo does not carry it and it was
+     *  left alone. */
+    private boolean neutralize(Path root, String worktreePath, WorktreeHostileConfig cfg) {
         Path target = root.resolve(cfg.file());
         if (!Files.exists(target) && !cfg.createIfAbsent()) {
-            log.debug("{} absent in the worktree — skipping (repo does not carry it)", cfg.file());
-            return;
+            return false;
         }
         try {
             Files.writeString(target, cfg.stub());
@@ -449,7 +505,7 @@ public final class GitWorktrees implements Worktrees {
         if (isTracked(root, cfg.file())) {
             exec("git", "-C", worktreePath, "update-index", "--skip-worktree", cfg.file());
         }
-        log.debug("neutralized {} — worker tool surface is launcher-mounted only", cfg.file());
+        return true;
     }
 
     @Override
@@ -486,25 +542,46 @@ public final class GitWorktrees implements Worktrees {
         }
         Path srcRoot = Path.of(repoRoot).toAbsolutePath().normalize();
         Path dstRoot = Path.of(worktreePath).toAbsolutePath().normalize();
+        List<String> copied = new ArrayList<>();
+        List<String> skipped = new ArrayList<>();
+        List<String> neutralized = new ArrayList<>();
         for (String rel : overlay) {
             Path src = srcRoot.resolve(rel).normalize();
             if (!Files.exists(src)) {
-                log.debug("parity overlay source missing — skipping {}", rel);
+                skipped.add(rel + " absent");
                 continue;
             }
             Path dst = dstRoot.resolve(rel).normalize();
             try {
                 Files.createDirectories(dst.getParent());
                 Files.copy(src, dst, StandardCopyOption.REPLACE_EXISTING, StandardCopyOption.COPY_ATTRIBUTES);
-                log.debug("copied parity overlay {}", rel);
+                copied.add(rel);
             } catch (IOException e) {
                 throw new WorktreeException("cannot copy overlay " + rel + ": " + e.getMessage(), e);
             }
             if (isTracked(dstRoot, rel)) {
                 exec("git", "-C", worktreePath, "update-index", "--skip-worktree", rel);
-                log.debug("marked overlay --skip-worktree {}", rel);
+                neutralized.add(rel);
             }
         }
+        // CB-148 point 3: a bare count ("copied 1") hides which candidates were even considered — the
+        // same shape of under-reporting this repo has been bitten by before. Name the denominator
+        // (every configured candidate), what was actually copied, and — for anything not copied —
+        // why, so a spawn's overlay outcome is legible from the log alone, no filesystem dig required.
+        String detail = copied.isEmpty() ? String.join(", ", skipped)
+                : skipped.isEmpty() ? String.join(", ", copied)
+                : String.join(", ", copied) + " (" + String.join(", ", skipped) + ")";
+        log.info("parity overlay: copied {} of {} candidates: {}", copied.size(), overlay.size(), detail);
+        // CB-134: a neutralized tracked file is otherwise a silent trap — a worker edits it, git
+        // ignores the change with no error, and nothing anywhere said the file could not be
+        // committed from this worktree. Name every file marked --skip-worktree here, with the
+        // consequence stated in the message itself, rather than adding a worktree-local marker
+        // file: acceptance criterion 1 requires the worktree hold exactly the configured overlay
+        // set and nothing else, so an extra marker would itself violate the fix.
+        if (!neutralized.isEmpty()) {
+            log.info("parity overlay marked --skip-worktree (cannot be committed from this worktree): {}",
+                    String.join(", ", neutralized));
+        }
     }
 
     @Override

fleetd/src/test/java/dev/ltms/fleet/FleetdBackendErrorPatternLookupTest.java

@@ -0,0 +1,59 @@
+package dev.ltms.fleet;
+
+import dev.ltms.fleet.inject.BackendErrorPatternLookup;
+import dev.ltms.fleet.peer.MemberRole;
+import dev.ltms.fleet.session.MemberSession;
+import org.junit.jupiter.api.DisplayName;
+import org.junit.jupiter.api.Test;
+
+import java.util.List;
+import java.util.Map;
+import java.util.regex.Pattern;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertNull;
+
+/**
+ * fleetd #248 / fleetd#201 Unit 5: {@link Fleetd#backendErrorPatternLookup} is the factory that
+ * replaced the local lambda {@code Fleetd.main} used to build {@code backendErrorPatterns} — one
+ * of the two arguments {@code CompletionResolver} lost cleanly (0 compile errors, every test still
+ * green) when this ticket's measurement dropped it alongside {@code backendErrorSink}. This class
+ * proves the factory's own behaviour; {@code FleetdCompletionResolverWiringTest} proves {@code
+ * main} still passes its result into {@code CompletionResolver}.
+ */
+class FleetdBackendErrorPatternLookupTest {
+
+    private static MemberSession session(String terminal, String profile) {
+        return new MemberSession("pane-" + terminal, terminal, profile, MemberRole.DEV,
+                "/cwd", null, 0L, 0L, 0, MemberSession.State.READY, null, null);
+    }
+
+    @Test
+    @DisplayName("a target on a profile with a configured pattern resolves to that pattern")
+    void configuredProfileResolves() {
+        Map<String, Pattern> byProfile = Map.of("terra", Pattern.compile("(?i)503"));
+        BackendErrorPatternLookup lookup =
+                Fleetd.backendErrorPatternLookup(() -> List.of(session("term1", "terra")), byProfile);
+
+        assertEquals("(?i)503", lookup.patternFor("term1").pattern());
+    }
+
+    @Test
+    @DisplayName("a target on a profile with no configured pattern resolves to null")
+    void unconfiguredProfileResolvesToNull() {
+        Map<String, Pattern> byProfile = Map.of("terra", Pattern.compile("x"));
+        BackendErrorPatternLookup lookup =
+                Fleetd.backendErrorPatternLookup(() -> List.of(session("term1", "sol")), byProfile);
+
+        assertNull(lookup.patternFor("term1"));
+    }
+
+    @Test
+    @DisplayName("an unknown target resolves to null")
+    void unknownTargetResolvesToNull() {
+        BackendErrorPatternLookup lookup =
+                Fleetd.backendErrorPatternLookup(List::of, Map.of("terra", Pattern.compile("x")));
+
+        assertNull(lookup.patternFor("term_stranger"));
+    }
+}

fleetd/src/test/java/dev/ltms/fleet/FleetdBackendErrorSinkTest.java

@@ -0,0 +1,229 @@
+package dev.ltms.fleet;
+
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import dev.ltms.fleet.config.FleetConfig;
+import dev.ltms.fleet.guard.SubscriptionGuard;
+import dev.ltms.fleet.herdr.AgentControl;
+import dev.ltms.fleet.herdr.FakeHerdr;
+import dev.ltms.fleet.herdr.HerdrClient;
+import dev.ltms.fleet.herdr.WorkspaceControl;
+import dev.ltms.fleet.inject.BackendErrorSink;
+import dev.ltms.fleet.member.ClaudeCodeLauncher;
+import dev.ltms.fleet.member.CompositePeerLauncher;
+import dev.ltms.fleet.mcp.PrimaryRegistry;
+import dev.ltms.fleet.msg.InMemoryReplyInbox;
+import dev.ltms.fleet.msg.ReplyPushLoop;
+import dev.ltms.fleet.peer.Capability;
+import dev.ltms.fleet.peer.PeerHandle;
+import dev.ltms.fleet.peer.PeerLauncher;
+import dev.ltms.fleet.peer.SpawnRequest;
+import dev.ltms.fleet.placement.BackendOutagePolicy;
+import dev.ltms.fleet.placement.BackendQuarantine;
+import dev.ltms.fleet.placement.PlacementPolicies;
+import dev.ltms.fleet.session.MemberSession;
+import dev.ltms.fleet.session.SessionManager;
+import org.junit.jupiter.api.AfterEach;
+import org.junit.jupiter.api.DisplayName;
+import org.junit.jupiter.api.Test;
+
+import java.util.ArrayList;
+import java.util.LinkedHashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.Set;
+import java.util.concurrent.CopyOnWriteArrayList;
+import java.util.concurrent.CountDownLatch;
+import java.util.concurrent.Executors;
+import java.util.concurrent.ScheduledExecutorService;
+import java.util.concurrent.TimeUnit;
+import java.util.concurrent.atomic.AtomicLong;
+import java.util.concurrent.atomic.AtomicReference;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertTrue;
+
+/**
+ * fleetd #248 / fleetd#201 Unit 5: {@link Fleetd#backendErrorSink} is the factory that replaced
+ * the local lambda {@code Fleetd.main} used to build {@code backendErrorSink} — the other half of
+ * the pair this ticket's measurement dropped cleanly (0 compile errors, every test still green).
+ *
+ * <p>Before this ticket, the closest thing to coverage was {@code
+ * dev.ltms.fleet.inject.BackendOutageFlowTest}, whose own class doc said it "mirrors {@code
+ * Fleetd.main}'s {@code backendErrorSink} lambda line-for-line" — a hand-copy that proves itself,
+ * never that {@code main} still wires the real thing. This class exercises the actual production
+ * factory instead. {@code FleetdCompletionResolverWiringTest} proves {@code main} still passes its
+ * result into {@code CompletionResolver}.
+ */
+class FleetdBackendErrorSinkTest {
+
+    private final List<ScheduledExecutorService> schedulers = new ArrayList<>();
+
+    @AfterEach
+    void tearDown() {
+        schedulers.forEach(ScheduledExecutorService::shutdownNow);
+    }
+
+    private static FleetConfig.Profile stubWorker(String profile, String credentialId) {
+        return new FleetConfig.Profile(profile, "http://gx00.gw:8000", "coder",
+                null, "FLEETD_WORKER_TOKEN", List.of("claude"), "tab", "fleetd-workers",
+                "w #{n}", null, null, null, null, null, null, null, null, null,
+                null, null, credentialId, null);
+    }
+
+    private static Map<String, FleetConfig.Profile> orderedProfiles() {
+        Map<String, FleetConfig.Profile> m = new LinkedHashMap<>();
+        m.put("terra", stubWorker("terra", "shared-openai"));
+        m.put("sol", stubWorker("sol", "shared-openai"));
+        return m;
+    }
+
+    /** Minimal recording {@code HerdrClient} for the LEAD pane — mirrors ReplyPushLoopTest's own. */
+    private static final class RecordingLeadClient implements HerdrClient {
+        private static final ObjectMapper MAPPER = new ObjectMapper();
+        private final List<Object> prompts = new CopyOnWriteArrayList<>();
+        volatile CountDownLatch sendLatch = new CountDownLatch(1);
+
+        @Override
+        public JsonNode call(String method, Object params) {
+            if ("agent.get".equals(method)) {
+                return MAPPER.createObjectNode().set("agent", MAPPER.createObjectNode()
+                        .put("terminal_id", "term_primary").put("agent_status", "idle"));
+            }
+            if ("agent.prompt".equals(method)) {
+                prompts.add(params);
+                sendLatch.countDown();
+            }
+            return MAPPER.createObjectNode();
+        }
+
+        @Override
+        public void close() {
+        }
+
+        int sendCount() {
+            return prompts.size();
+        }
+    }
+
+    /** A {@link PeerLauncher} that never actually spawns — enough to construct a bare {@link SessionManager}. */
+    private static final class NeverSpawnsLauncher implements PeerLauncher {
+        @Override
+        public Set<Capability> capabilities() {
+            return Set.of();
+        }
+
+        @Override
+        public Set<Capability> capabilitiesFor(String profileName) {
+            return Set.of();
+        }
+
+        @Override
+        public PeerHandle spawn(SpawnRequest req) {
+            throw new UnsupportedOperationException("not reachable — this test never acquires a session");
+        }
+
+        @Override
+        public Set<String> profiles() {
+            return Set.of();
+        }
+
+        @Override
+        public String defaultProfile() {
+            return null;
+        }
+
+        @Override
+        public String effectiveCwd(SpawnRequest req) {
+            throw new UnsupportedOperationException("not reachable — this test never acquires a session");
+        }
+
+        @Override
+        public List<String> parityOverlay(String profileName) {
+            return List.of();
+        }
+
+        @Override
+        public List<?> list() {
+            return List.of();
+        }
+
+        @Override
+        public int reapOrphanWorkers() {
+            return 0;
+        }
+
+        @Override
+        public void stop(String id) {
+        }
+
+        @Override
+        public boolean clearContext(String id) {
+            return false;
+        }
+    }
+
+    @Test
+    @DisplayName("a target with no resolvable profile logs and returns without recording an incident (never throws)")
+    void unresolvableProfileDoesNotRecordOrThrow() {
+        SessionManager sessions = new SessionManager(new NeverSpawnsLauncher());
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        RecordingLeadClient leadClient = new RecordingLeadClient();
+        InMemoryReplyInbox inbox = new InMemoryReplyInbox();
+        PrimaryRegistry registry = new PrimaryRegistry(null);
+        ScheduledExecutorService scheduler = Executors.newSingleThreadScheduledExecutor();
+        schedulers.add(scheduler);
+        ReplyPushLoop pushLoop = new ReplyPushLoop(registry, new AgentControl(leadClient), inbox, scheduler, 3, 50);
+
+        BackendErrorSink sink = Fleetd.backendErrorSink(sessions, Map::of, outagePolicy, () -> pushLoop);
+        sink.onBackendError("term_unmapped", "matched line", "503 Service Unavailable");
+
+        assertTrue(outagePolicy.remainingCoolOffSeconds("shared-openai").isEmpty(),
+                "no credential is ever resolvable here, so nothing must be recorded");
+    }
+
+    @Test
+    @DisplayName("two distinct targets classified through the real factory start an incident and cool the credential")
+    void twoDistinctTargetsStartAnIncident() throws Exception {
+        FakeHerdr herdr = new FakeHerdr()
+                .readText("⏺ 503 Service Unavailable: upstream credential rejected\n❯ ");
+        Map<String, FleetConfig.Profile> profiles = orderedProfiles();
+        AtomicLong clockNanos = new AtomicLong(0L);
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(clockNanos::get);
+        ClaudeCodeLauncher adapter = new ClaudeCodeLauncher(new AgentControl(herdr),
+                new WorkspaceControl(herdr), new SubscriptionGuard(Set.of("gx00.gw")),
+                profiles, "terra", _ -> "tok");
+        CompositePeerLauncher workers = new CompositePeerLauncher(List.of(adapter), "terra", profiles,
+                PlacementPolicies.weighted(), _ -> 0, null, BackendQuarantine.none(), outagePolicy);
+        SessionManager sessions = new SessionManager(workers);
+        MemberSession s1 = sessions.acquire("terra", null, null, null);
+        MemberSession s2 = sessions.acquire("terra", null, null, null);
+
+        PrimaryRegistry registry = new PrimaryRegistry(null);
+        registry.recordDelegation(s1.terminalId(), "term_primary");
+        registry.recordDelegation(s2.terminalId(), "term_primary");
+        InMemoryReplyInbox inbox = new InMemoryReplyInbox();
+        inbox.own(s1.terminalId());
+        inbox.own(s2.terminalId());
+        RecordingLeadClient leadClient = new RecordingLeadClient();
+        ScheduledExecutorService scheduler = Executors.newSingleThreadScheduledExecutor();
+        schedulers.add(scheduler);
+        ReplyPushLoop pushLoop = new ReplyPushLoop(registry, new AgentControl(leadClient), inbox, scheduler, 3, 50);
+        AtomicReference<ReplyPushLoop> pushLoopRef = new AtomicReference<>(pushLoop);
+
+        // The exact object under test: Fleetd's real production factory, not a hand copy.
+        BackendErrorSink sink = Fleetd.backendErrorSink(sessions, () -> profiles, outagePolicy, pushLoopRef::get);
+
+        sink.onBackendError(s1.terminalId(), "matched line", "503 Service Unavailable");
+        assertTrue(outagePolicy.remainingCoolOffSeconds("shared-openai").isEmpty(),
+                "one distinct target must not start a cool-off");
+
+        sink.onBackendError(s2.terminalId(), "matched line", "503 Service Unavailable");
+
+        assertTrue(leadClient.sendLatch.await(3, TimeUnit.SECONDS),
+                "the second distinct target must cross the threshold and nudge the lead");
+        var remaining = outagePolicy.remainingCoolOffSeconds("shared-openai");
+        assertTrue(remaining.isPresent(), "two distinct targets must start a cool-off");
+        assertEquals(1, leadClient.sendCount());
+    }
+}

fleetd/src/test/java/dev/ltms/fleet/FleetdCompletionResolverWiringTest.java

@@ -0,0 +1,89 @@
+package dev.ltms.fleet;
+
+import java.nio.file.Files;
+import java.nio.file.Path;
+import org.junit.jupiter.api.DisplayName;
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.assertFalse;
+import static org.junit.jupiter.api.Assertions.assertTrue;
+
+/**
+ * fleetd #248: this is the test that was actually missing. {@code Fleetd.main} builds its {@code
+ * CompletionResolver} from an 8-argument constructor, and the ticket's own measurement proved two
+ * ways to silently unwire it — both compiled with 0 errors and left every existing test green:
+ *
+ * <ul>
+ *   <li>replacing the worktree/branch argument (the 8th) with {@code _ -> null} — drops
+ *       fleetd#241's fallback-report location entirely;</li>
+ *   <li>replacing {@code backendErrorPatterns, backendErrorSink} (5th/6th) with {@code
+ *       BackendErrorPatternLookup.legacy(), BackendErrorSink.none()} — drops fleetd#201 Unit 5's
+ *       backend-error classification and cool-off entirely.</li>
+ * </ul>
+ *
+ * <p>Neither mutation could be caught by any test that constructs its own {@code
+ * CompletionResolver} (every test before this one did exactly that) or by a test of {@link
+ * Fleetd#worktreeBranchLookup}, {@link Fleetd#backendErrorPatternLookup}, or {@link
+ * Fleetd#backendErrorSink} in isolation (see {@code FleetdWorktreeBranchLookupTest}, {@code
+ * FleetdBackendErrorPatternLookupTest}, {@code FleetdBackendErrorSinkTest}) — those prove the
+ * factories work, never that {@code main} still calls them. This class is a plain source-text
+ * assertion on {@code Fleetd.java} — crude, but honest about what it checks, and it turns red the
+ * instant the wiring is dropped, mirroring the same fallback shape {@link
+ * FleetdFleetAppConstructionTest} already uses for a different constructor argument.
+ *
+ * <p><b>This test checks source text, not runtime behaviour.</b> It never constructs a {@code
+ * CompletionResolver} and never runs {@code main}.
+ */
+class FleetdCompletionResolverWiringTest {
+
+    private static String fleetdSource() throws Exception {
+        return Files.readString(Path.of("src/main/java/dev/ltms/fleet/Fleetd.java"));
+    }
+
+    @Test
+    @DisplayName("[SOURCE TEXT] CompletionResolver's construction call still names backendErrorPatterns and backendErrorSink")
+    void backendErrorArgumentsAreStillNamedAtTheCallSite() throws Exception {
+        String source = fleetdSource();
+        assertTrue(source.contains(
+                        "exhaustionSink, backendErrorPatterns, backendErrorSink, System::nanoTime,"),
+                "CompletionResolver's construction call must still pass backendErrorPatterns and "
+                        + "backendErrorSink as its 5th/6th arguments. Replacing them with "
+                        + "BackendErrorPatternLookup.legacy()/BackendErrorSink.none() (fleetd #248's measured "
+                        + "mutation) compiles with 0 errors and leaves every behavioural test green — this "
+                        + "source check is what must go red instead.");
+    }
+
+    @Test
+    @DisplayName("[SOURCE TEXT] CompletionResolver's construction call still passes worktreeBranchLookup(sessions::roster)")
+    void worktreeBranchLookupIsStillPassedAtTheCallSite() throws Exception {
+        String source = fleetdSource();
+        assertTrue(source.contains("worktreeBranchLookup(sessions::roster)"),
+                "CompletionResolver's construction call must still pass worktreeBranchLookup(sessions::roster) "
+                        + "as its 8th (last) argument. Replacing it with the inert `_ -> null` (fleetd #248's "
+                        + "other measured mutation) compiles with 0 errors and leaves every behavioural test "
+                        + "green — this source check is what must go red instead.");
+        assertFalse(source.contains("System::nanoTime,\n                _ -> null"),
+                "the worktree/branch argument must never regress to the inert `_ -> null` literal");
+    }
+
+    @Test
+    @DisplayName("[SOURCE TEXT] backendErrorPatterns is assigned from the extracted backendErrorPatternLookup(...) factory")
+    void backendErrorPatternsComesFromTheFactory() throws Exception {
+        String source = fleetdSource();
+        assertTrue(source.contains(
+                        "BackendErrorPatternLookup backendErrorPatterns = backendErrorPatternLookup(sessions::roster,"),
+                "backendErrorPatterns must be assigned from Fleetd.backendErrorPatternLookup(...), not an "
+                        + "inline lambda that a source check on the CompletionResolver call alone cannot see "
+                        + "through");
+    }
+
+    @Test
+    @DisplayName("[SOURCE TEXT] backendErrorSink is assigned from the extracted backendErrorSink(...) factory")
+    void backendErrorSinkComesFromTheFactory() throws Exception {
+        String source = fleetdSource();
+        assertTrue(source.contains(
+                        "BackendErrorSink backendErrorSink = backendErrorSink(sessions, () -> config.get().profiles(),"),
+                "backendErrorSink must be assigned from Fleetd.backendErrorSink(...), not an inline lambda "
+                        + "that a source check on the CompletionResolver call alone cannot see through");
+    }
+}

fleetd/src/test/java/dev/ltms/fleet/FleetdWorktreeBranchLookupTest.java

@@ -0,0 +1,64 @@
+package dev.ltms.fleet;
+
+import dev.ltms.fleet.inject.CompletionResolver;
+import dev.ltms.fleet.peer.MemberRole;
+import dev.ltms.fleet.session.MemberSession;
+import org.junit.jupiter.api.DisplayName;
+import org.junit.jupiter.api.Test;
+
+import java.util.ArrayList;
+import java.util.List;
+import java.util.function.Function;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertNull;
+
+/**
+ * fleetd #248: {@link Fleetd#worktreeBranchLookup} is the factory that replaced the anonymous
+ * lambda {@code Fleetd.main} used to build inline, as the 8th (last) argument to {@code
+ * CompletionResolver}'s constructor. Before this ticket that argument was untestable wiring:
+ * replacing it with {@code _ -> null} compiled clean and every existing test stayed green, because
+ * every existing test builds its own {@code CompletionResolver} directly rather than going through
+ * {@code main}. This class proves the factory's own behaviour; {@code
+ * FleetdCompletionResolverWiringTest} proves {@code main} still passes it in.
+ */
+class FleetdWorktreeBranchLookupTest {
+
+    private static MemberSession session(String terminal, String worktree, String branch) {
+        return new MemberSession("pane-" + terminal, terminal, "terra", MemberRole.DEV,
+                "/cwd", null, 0L, 0L, 0, MemberSession.State.READY, worktree, branch);
+    }
+
+    @Test
+    @DisplayName("a known target resolves to its session's worktree and branch")
+    void knownTargetResolves() {
+        Function<String, CompletionResolver.WorktreeBranch> lookup =
+                Fleetd.worktreeBranchLookup(() -> List.of(session("term1", "/wt/worker_x", "worker/x")));
+
+        CompletionResolver.WorktreeBranch resolved = lookup.apply("term1");
+
+        assertEquals("/wt/worker_x", resolved.worktree());
+        assertEquals("worker/x", resolved.branch());
+    }
+
+    @Test
+    @DisplayName("an unknown target resolves to null, not a thrown exception")
+    void unknownTargetResolvesToNull() {
+        Function<String, CompletionResolver.WorktreeBranch> lookup =
+                Fleetd.worktreeBranchLookup(() -> List.of(session("term1", "/wt/worker_x", "worker/x")));
+
+        assertNull(lookup.apply("term_stranger"));
+    }
+
+    @Test
+    @DisplayName("the roster is read through the supplier on every call, not snapshotted")
+    void rosterIsReadThroughOnEveryCall() {
+        List<MemberSession> roster = new ArrayList<>();
+        Function<String, CompletionResolver.WorktreeBranch> lookup = Fleetd.worktreeBranchLookup(() -> roster);
+
+        assertNull(lookup.apply("term_late"));
+        roster.add(session("term_late", "/wt/late", "worker/late"));
+
+        assertEquals("/wt/late", lookup.apply("term_late").worktree());
+    }
+}

fleetd/src/test/java/dev/ltms/fleet/config/ConfigRefTest.java

@@ -386,6 +386,54 @@ class ConfigRefTest {
         assertEquals("rate limit exceeded", ref.get().profiles().get("sonnet").exhaustedPattern());
     }
 
+    /**
+     * fleetd #201 Unit 5: errorPattern is compiled once into Fleetd.main's backend-error pattern
+     * map at startup (see BackendErrorPatternLookup), the same way exhaustedPattern is above — a
+     * reload never re-reads it either, so a changed value must be reported deferred.
+     */
+    @Test
+    void changingAProfilesErrorPatternIsReportedAsDeferred(@TempDir Path dir) throws Exception {
+        Path f = dir.resolve("fleetd.yaml");
+        Files.writeString(f, """
+                bind:
+                  host: 127.0.0.1
+                  port: 8765
+                herdrSocket: ~/.config/herdr/herdr.sock
+                profiles:
+                  sonnet:
+                    baseUrl: http://gx00.gw:8000
+                    model: sonnet
+                    errorPattern: "credential outage"
+                guard:
+                  offSubscriptionHosts:
+                    - gx00.gw
+                """);
+        ConfigRef ref = refFor(f);
+
+        Files.writeString(f, """
+                bind:
+                  host: 127.0.0.1
+                  port: 8765
+                herdrSocket: ~/.config/herdr/herdr.sock
+                profiles:
+                  sonnet:
+                    baseUrl: http://gx00.gw:8000
+                    model: sonnet
+                    errorPattern: "provider 5xx"
+                guard:
+                  offSubscriptionHosts:
+                    - gx00.gw
+                """);
+        ConfigRef.Outcome out = ref.reload();
+
+        assertTrue(out.applied());
+        assertEquals(1, out.deferred().size(), out.deferred().toString());
+        assertTrue(out.deferred().getFirst().contains("sonnet"), out.deferred().toString());
+        assertTrue(out.deferred().getFirst().contains("launch settings"), out.deferred().toString());
+        // The snapshot still carries the new value — a restart is what makes it take effect.
+        assertEquals("provider 5xx", ref.get().profiles().get("sonnet").errorPattern());
+    }
+
     @Test
     void aFixedRefHasNoFileAndRefusesToReload() {
         FleetConfig cfg = new FleetConfig(null, null, null, null, null, null,

fleetd/src/test/java/dev/ltms/fleet/config/FleetConfigTest.java

@@ -86,6 +86,85 @@ class FleetConfigTest {
                 "unset means off — today's behaviour, unchanged");
     }
 
+    // ── fleetd #201 Unit 5: errorPattern ────────────────────────────────────────────────────────
+
+    @Test
+    void aProfileWithAnErrorPatternBindsAndHasErrorPatternIsTrue(@TempDir Path dir) throws Exception {
+        Path f = dir.resolve("error-pattern.yaml");
+        Files.writeString(f, """
+                profiles:
+                  ltms-local:
+                    baseUrl: http://gx00.gw:8000
+                    errorPattern: "(?i)\\\\bcredential outage\\\\b"
+                """);
+
+        FleetConfig.Profile w = FleetConfig.load(f).profiles().get("ltms-local");
+        assertTrue(w.hasErrorPattern(), "errorPattern binds and enables fleetd #201 Unit 5 classification");
+        assertEquals("(?i)\\bcredential outage\\b", w.errorPattern());
+    }
+
+    @Test
+    void aProfileWithNoErrorPatternLeavesItNullAndHasErrorPatternIsFalse(@TempDir Path dir) throws Exception {
+        Path f = dir.resolve("no-error-pattern.yaml");
+        Files.writeString(f, """
+                profiles:
+                  ltms-local:
+                    baseUrl: http://gx00.gw:8000
+                """);
+
+        FleetConfig.Profile w = FleetConfig.load(f).profiles().get("ltms-local");
+        assertNull(w.errorPattern(), "unset means 'use the built-in fallback' — never off");
+        assertFalse(w.hasErrorPattern());
+    }
+
+    @Test
+    void aBlankErrorPatternNormalizesToNullJustLikeUnset(@TempDir Path dir) throws Exception {
+        Path f = dir.resolve("blank-error-pattern.yaml");
+        Files.writeString(f, """
+                profiles:
+                  ltms-local:
+                    baseUrl: http://gx00.gw:8000
+                    errorPattern: "   "
+                """);
+
+        FleetConfig.Profile w = FleetConfig.load(f).profiles().get("ltms-local");
+        assertNull(w.errorPattern());
+        assertFalse(w.hasErrorPattern());
+    }
+
+    @Test
+    void aProfileWithAMalformedErrorPatternIsRejectedAtLoadNamingTheProfileAndKey(@TempDir Path dir)
+            throws Exception {
+        Path f = dir.resolve("malformed-error-pattern.yaml");
+        Files.writeString(f, """
+                profiles:
+                  ltms-local:
+                    baseUrl: http://gx00.gw:8000
+                    errorPattern: "(unterminated["
+                """);
+
+        IllegalStateException e = assertThrows(IllegalStateException.class, () -> FleetConfig.load(f));
+        assertTrue(e.getMessage().contains("ltms-local"), "the offending profile is named: " + e.getMessage());
+        assertTrue(e.getMessage().contains("errorPattern"), "the offending key is named: " + e.getMessage());
+    }
+
+    @Test
+    void withProfileCarriesErrorPatternThrough(@TempDir Path dir) throws Exception {
+        Path f = dir.resolve("with-profile-error-pattern.yaml");
+        Files.writeString(f, """
+                profiles:
+                  ltms-local:
+                    baseUrl: http://gx00.gw:8000
+                    errorPattern: "credential outage"
+                """);
+
+        // The compact constructor defaults `profile` to the map key via withProfile(name) — see
+        // loadsFullConfig above. errorPattern must survive that same rebuild.
+        FleetConfig.Profile w = FleetConfig.load(f).profiles().get("ltms-local");
+        assertEquals("ltms-local", w.profile());
+        assertEquals("credential outage", w.errorPattern());
+    }
+
     @Test
     void appliesDefaultsForMissingSections(@TempDir Path dir) throws Exception {
         Path f = dir.resolve("minimal.yaml");
@@ -1313,6 +1392,7 @@ class FleetConfigTest {
                     weight: 0.5
                     maxLoad: 2
                     exhaustedPattern: "usage limit has been reached"
+                    errorPattern: "credential outage"
                 placement: weighted
                 lifecycle:
                   idleTtlSeconds: 300
@@ -1350,6 +1430,8 @@ class FleetConfigTest {
         assertEquals(2, w.maxLoad(), "maxLoad binds as an integer");
         assertTrue(w.hasExhaustedPattern(), "exhaustedPattern binds and enables the CB-578 stage A classification");
         assertEquals("usage limit has been reached", w.exhaustedPattern());
+        assertTrue(w.hasErrorPattern(), "errorPattern binds and enables fleetd #201 Unit 5 classification");
+        assertEquals("credential outage", w.errorPattern());
 
         assertEquals("weighted", cfg.placement(), "placement binds at the top level");
         assertEquals(300, cfg.lifecycle().idleTtlSeconds());
@@ -1781,7 +1863,7 @@ class FleetConfigTest {
     }
 
     @Test
-    void parityOverlayDefaultsToEnvFilesOnly(@TempDir Path dir) throws Exception {
+    void parityOverlayDefaultsToDotEnvOnly(@TempDir Path dir) throws Exception {
         Path f = dir.resolve("no-overlay.yaml");
         Files.writeString(f, """
                 bind:
@@ -1792,9 +1874,11 @@ class FleetConfigTest {
                 """);
 
         FleetConfig cfg = FleetConfig.load(f);
-        assertEquals(List.of(".env", ".envrc"),
+        assertEquals(List.of(".env"),
                 cfg.profiles().get("gx10").parityOverlay(),
-                "the default parity overlay is the env files; settings.local.json is no longer copied by default");
+                "the default parity overlay is .env only (CB-148): .envrc is executable shell that "
+                        + "direnv runs on every cd, so it is no longer copied by default; settings.local.json "
+                        + "is also not copied by default");
     }
 
     @Test
@@ -1815,6 +1899,25 @@ class FleetConfigTest {
                 "an operator's explicit list survives verbatim — the default only changes when unset");
     }
 
+    @Test
+    void parityOverlayExplicitEnvrcOptInStillWorks(@TempDir Path dir) throws Exception {
+        Path f = dir.resolve("explicit-envrc-overlay.yaml");
+        Files.writeString(f, """
+                bind:
+                  port: 8080
+                profiles:
+                  gx10:
+                    baseUrl: http://gx10.gw:8000
+                    parityOverlay: [".env", ".envrc"]
+                """);
+
+        FleetConfig cfg = FleetConfig.load(f);
+        assertEquals(List.of(".env", ".envrc"),
+                cfg.profiles().get("gx10").parityOverlay(),
+                "an operator can still opt into copying .envrc explicitly (CB-148); it is only "
+                        + "dropped from the unset default, not removed as a capability");
+    }
+
     // ── CB-542: subscription:true must not smuggle an unguarded endpoint via env: ───────────────
 
     @Test

fleetd/src/test/java/dev/ltms/fleet/herdr/FakeHerdr.java

@@ -49,6 +49,7 @@ public final class FakeHerdr implements HerdrClient {
     private int pinnedStarts = 0;   // how many upcoming agent.start calls report a fixed pane
     private String pinnedStartTerminal;
     private String pinnedStartPane;
+    private Runnable onAgentStart; // fires the instant agent.start is called — see onAgentStart(Runnable)
     private volatile int agentGetOkCalls = Integer.MAX_VALUE; // how many agent.get calls succeed first
     private volatile String agentGetFailCode = null; // error code every agent.get call after that reports
 
@@ -152,6 +153,18 @@ public final class FakeHerdr implements HerdrClient {
         return this;
     }
 
+    /**
+     * Run {@code hook} synchronously the instant an {@code agent.start} call reaches this fake —
+     * i.e. the instant the peer PROCESS would start against a real herdr daemon. A test uses this
+     * to assert something is already true at that exact point (rather than merely true once
+     * {@code spawn()} returns), e.g. fleetd #149's trust-dialog seed having already been written to
+     * disk before the process herdr would launch ever starts.
+     */
+    public FakeHerdr onAgentStart(Runnable hook) {
+        this.onAgentStart = hook;
+        return this;
+    }
+
 
     /**
      * Seed a named agent into {@code agent.list} (e.g. an orphaned worker for CB-117 reaper tests).
@@ -250,6 +263,9 @@ public final class FakeHerdr implements HerdrClient {
                 case "agent.read" -> mapper.readTree(mapper.writeValueAsString(
                         java.util.Map.of("type", "agent_read", "read", java.util.Map.of("text", readText))));
                 case "agent.start" -> {
+                    if (onAgentStart != null) {
+                        onAgentStart.run();
+                    }
                     // Protocol 19: kind and pane_id are required — reject like the real daemon.
                     java.util.Map<?, ?> p = params instanceof java.util.Map<?, ?> m ? m : java.util.Map.of();
                     for (String required : new String[]{"kind", "pane_id"}) {

fleetd/src/test/java/dev/ltms/fleet/inject/BackendOutageFlowTest.java

@@ -0,0 +1,272 @@
+package dev.ltms.fleet.inject;
+
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import dev.ltms.fleet.Fleetd;
+import dev.ltms.fleet.config.FleetConfig;
+import dev.ltms.fleet.guard.SubscriptionGuard;
+import dev.ltms.fleet.herdr.AgentControl;
+import dev.ltms.fleet.herdr.FakeHerdr;
+import dev.ltms.fleet.herdr.HerdrClient;
+import dev.ltms.fleet.herdr.WorkspaceControl;
+import dev.ltms.fleet.mcp.PrimaryRegistry;
+import dev.ltms.fleet.member.ClaudeCodeLauncher;
+import dev.ltms.fleet.member.CompositePeerLauncher;
+import dev.ltms.fleet.msg.InMemoryReplyInbox;
+import dev.ltms.fleet.msg.Rendezvous;
+import dev.ltms.fleet.msg.ReplyPushLoop;
+import dev.ltms.fleet.peer.SpawnRequest;
+import dev.ltms.fleet.placement.BackendOutagePolicy;
+import dev.ltms.fleet.placement.BackendQuarantine;
+import dev.ltms.fleet.placement.PlacementException;
+import dev.ltms.fleet.placement.PlacementPolicies;
+import dev.ltms.fleet.session.MemberSession;
+import dev.ltms.fleet.session.SessionManager;
+import org.junit.jupiter.api.AfterEach;
+import org.junit.jupiter.api.Test;
+
+import java.util.ArrayList;
+import java.util.LinkedHashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.Optional;
+import java.util.Set;
+import java.util.concurrent.CopyOnWriteArrayList;
+import java.util.concurrent.CountDownLatch;
+import java.util.concurrent.Executors;
+import java.util.concurrent.ScheduledExecutorService;
+import java.util.concurrent.TimeUnit;
+import java.util.concurrent.atomic.AtomicLong;
+import java.util.concurrent.atomic.AtomicReference;
+import java.util.regex.Pattern;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+/**
+ * fleetd #201 / #227 Unit 5: proves the WIRED production path — a real {@link CompletionResolver}
+ * pane-scrape classification, through the exact {@code BackendErrorSink} ordering {@code Fleetd.main}
+ * builds (mirrored here line for line since the lambda itself lives inline in {@code Fleetd.main}),
+ * into a real {@link BackendOutagePolicy}, a real {@link CompositePeerLauncher} spawn gate, and a
+ * real {@link ReplyPushLoop} lead nudge — never by calling {@link BackendOutagePolicy#record} or the
+ * sink directly, which is what the finer-grained unit tests in
+ * {@code dev.ltms.fleet.member.CompositePeerLauncherTest} and {@code ReplyPushLoopTest} do for their
+ * own concerns. This class exists to catch a wiring mistake those unit tests cannot see: each of them
+ * hands the collaborator a value it assumes was computed correctly one layer up.
+ *
+ * <p>{@code fleet_list}/{@code fleet_profiles} rendering of the resulting cool-off (criteria 11/12)
+ * is proven separately in {@code dev.ltms.fleet.mcp.FleetMcpTest} — {@code FleetMcp}'s view methods
+ * are package-private to {@code dev.ltms.fleet.mcp} and this class needs {@code CompletionResolver}'s
+ * package-private {@code InFlight}, so the two cannot share one test class. What matters for THIS
+ * test is only that the same {@link BackendOutagePolicy} object the real scrape path populates is the
+ * one {@code FleetMcp} reads from — proven by asserting directly against
+ * {@link BackendOutagePolicy#remainingCoolOffSeconds} after the real classification below.
+ */
+class BackendOutageFlowTest {
+
+    private final List<ScheduledExecutorService> schedulers = new ArrayList<>();
+
+    @AfterEach
+    void tearDown() {
+        schedulers.forEach(ScheduledExecutorService::shutdownNow);
+    }
+
+    private static FleetConfig.Profile stubWorker(String profile, String credentialId) {
+        return new FleetConfig.Profile(profile, "http://gx00.gw:8000", "coder",
+                null, "FLEETD_WORKER_TOKEN", List.of("claude"), "tab", "fleetd-workers",
+                "w #{n}", null, null, null, null, null, null, null, null, null,
+                null, null, credentialId, null);
+    }
+
+    /** Minimal recording {@code HerdrClient} for the LEAD pane — mirrors ReplyPushLoopTest's own. */
+    private static final class RecordingLeadClient implements HerdrClient {
+        private static final ObjectMapper MAPPER = new ObjectMapper();
+        private final List<Object> prompts = new CopyOnWriteArrayList<>();
+        volatile CountDownLatch sendLatch = new CountDownLatch(1);
+
+        @Override
+        public JsonNode call(String method, Object params) {
+            if ("agent.get".equals(method)) {
+                return MAPPER.createObjectNode().set("agent", MAPPER.createObjectNode()
+                        .put("terminal_id", "term_primary").put("agent_status", "idle"));
+            }
+            if ("agent.prompt".equals(method)) {
+                prompts.add(params);
+                sendLatch.countDown();
+            }
+            return MAPPER.createObjectNode();
+        }
+
+        @Override
+        public void close() {
+        }
+
+        int sendCount() {
+            return prompts.size();
+        }
+
+        String lastText() {
+            return ((Map<?, ?>) prompts.get(prompts.size() - 1)).get("text").toString();
+        }
+    }
+
+    /** Every collaborator the flow wires together, built exactly once per test. */
+    private final class Flow {
+        final FakeHerdr herdr = new FakeHerdr()
+                .readText("⏺ 503 Service Unavailable: upstream credential rejected\n❯ ");
+        final Map<String, FleetConfig.Profile> profiles = orderedProfiles();
+        final AtomicLong clockNanos = new AtomicLong(0L);
+        final BackendOutagePolicy outagePolicy = new BackendOutagePolicy(clockNanos::get);
+        final CompositePeerLauncher workers;
+        final SessionManager sessions;
+        final MemberSession s1;
+        final MemberSession s2;
+        final PrimaryRegistry registry = new PrimaryRegistry(null);
+        final InMemoryReplyInbox inbox = new InMemoryReplyInbox();
+        final RecordingLeadClient leadClient = new RecordingLeadClient();
+        final ReplyPushLoop pushLoop;
+        final CompletionResolver resolver;
+        final Rendezvous rendezvous = new Rendezvous();
+
+        Flow() {
+            ClaudeCodeLauncher adapter = new ClaudeCodeLauncher(new AgentControl(herdr),
+                    new WorkspaceControl(herdr), new SubscriptionGuard(Set.of("gx00.gw")),
+                    profiles, "terra", _ -> "tok");
+            workers = new CompositePeerLauncher(List.of(adapter), "terra", profiles,
+                    PlacementPolicies.weighted(), _ -> 0, null, BackendQuarantine.none(), outagePolicy);
+            sessions = new SessionManager(workers);
+            // Two REAL member sessions on "terra" — the roster lookup the production sink does.
+            s1 = sessions.acquire("terra", null, null, null);
+            s2 = sessions.acquire("terra", null, null, null);
+            registry.recordDelegation(s1.terminalId(), "term_primary");
+            registry.recordDelegation(s2.terminalId(), "term_primary");
+            inbox.own(s1.terminalId());
+            inbox.own(s2.terminalId());
+
+            ScheduledExecutorService scheduler = Executors.newSingleThreadScheduledExecutor();
+            schedulers.add(scheduler);
+            pushLoop = new ReplyPushLoop(registry, new AgentControl(leadClient), inbox, scheduler, 3, 50);
+            AtomicReference<ReplyPushLoop> pushLoopRef = new AtomicReference<>(pushLoop);
+
+            // fleetd #248 follow-up: this used to be a 30-line hand-copy of Fleetd.main's
+            // backendErrorSink lambda, with a comment promising it mirrored production "EXACTLY".
+            // That promise is exactly the problem: a copy proves the copy. Editing or deleting the
+            // real sink left this whole flow test green, because it never touched the real sink.
+            // #248 made Fleetd.backendErrorSink public precisely so a cross-package test could
+            // drive the real object, so this now calls it. Every assertion below is about
+            // production code again.
+            BackendErrorSink backendErrorSink =
+                    Fleetd.backendErrorSink(sessions, () -> profiles, outagePolicy, pushLoopRef::get);
+
+            BackendErrorPatternLookup patterns = target -> Pattern.compile("(?i)503 Service Unavailable");
+            resolver = new CompletionResolver(new AgentControl(herdr), rendezvous, ExhaustedPatternLookup.none(),
+                    ExhaustionSink.none(), patterns, backendErrorSink);
+        }
+
+        /** Drives one real pane-scrape classification for {@code target} through {@code resolver}. */
+        void classifyBackendErrorOn(String target) {
+            var waiter = rendezvous.open(target);
+            resolver.resolve(target, new CompletionResolver.InFlight(waiter, null));
+            assertEquals(Rendezvous.Kind.FAILED, waiter.getNow(null).kind(),
+                    "a classified backend error still resolves the blocked send as a failure");
+        }
+    }
+
+    private static Map<String, FleetConfig.Profile> orderedProfiles() {
+        Map<String, FleetConfig.Profile> m = new LinkedHashMap<>();
+        m.put("terra", stubWorker("terra", "shared-openai"));
+        m.put("sol", stubWorker("sol", "shared-openai"));
+        return m;
+    }
+
+    private long agentStartCalls(FakeHerdr herdr) {
+        return herdr.calls.stream().filter(c -> c.method().equals("agent.start")).count();
+    }
+
+    // --- criteria 4/5: two DISTINCT targets start an incident; one target never does ------------
+
+    @Test
+    void aSingleClassifiedBackendErrorOnOneTargetNeverCoolsTheCredential() {
+        Flow flow = new Flow();
+
+        flow.classifyBackendErrorOn(flow.s1.terminalId());
+
+        assertTrue(flow.outagePolicy.remainingCoolOffSeconds("shared-openai").isEmpty(),
+                "one distinct target's classified error must not start a cool-off");
+        assertEquals(0, flow.leadClient.sendCount(), "no incident yet, so no lead nudge");
+        assertDoesNotThrow(() -> flow.workers.spawn(new SpawnRequest("terra", null, null)),
+                "terra must still be spawnable after only one target's error");
+    }
+
+    @Test
+    void twoDistinctTargetsClassifiedThroughTheRealScrapeStartAnIncidentAndCoolTheCredential() throws Exception {
+        Flow flow = new Flow();
+
+        flow.classifyBackendErrorOn(flow.s1.terminalId());
+        assertTrue(flow.outagePolicy.remainingCoolOffSeconds("shared-openai").isEmpty());
+
+        flow.classifyBackendErrorOn(flow.s2.terminalId());
+
+        assertTrue(flow.leadClient.sendLatch.await(3, TimeUnit.SECONDS),
+                "the second distinct target must cross the threshold and nudge the lead");
+        Thread.sleep(150);
+        var remaining = flow.outagePolicy.remainingCoolOffSeconds("shared-openai");
+        assertTrue(remaining.isPresent(), "two distinct targets must start a cool-off");
+        assertEquals(60L, remaining.getAsLong(),
+                "this is the exact fact FleetMcp.capacityView/profiles reads as coolingOffForSeconds "
+                        + "(see FleetMcpTest for the rendering side)");
+    }
+
+    // --- criterion 13: the incident reaches the lead through ReplyPushLoop exactly once ----------
+
+    @Test
+    void theIncidentSendsExactlyOneNudgeToTheLeadNamingBothCredentialAndTargets() throws Exception {
+        Flow flow = new Flow();
+
+        flow.classifyBackendErrorOn(flow.s1.terminalId());
+        flow.classifyBackendErrorOn(flow.s2.terminalId());
+
+        assertTrue(flow.leadClient.sendLatch.await(3, TimeUnit.SECONDS));
+        Thread.sleep(200);
+        assertEquals(1, flow.leadClient.sendCount(), "one incident must send exactly one nudge");
+        String nudge = flow.leadClient.lastText();
+        assertTrue(nudge.contains("shared-openai"), nudge);
+        assertTrue(nudge.contains(flow.s1.terminalId()) && nudge.contains(flow.s2.terminalId()), nudge);
+        assertTrue(nudge.contains("60"), nudge);
+    }
+
+    // --- criterion 6: explicit spawn is refused BEFORE the adapter is ever called -----------------
+
+    @Test
+    void explicitSpawnOnACooledCredentialIsRefusedBeforeReachingTheAdapter() throws Exception {
+        Flow flow = new Flow();
+        flow.classifyBackendErrorOn(flow.s1.terminalId());
+        flow.classifyBackendErrorOn(flow.s2.terminalId());
+        assertTrue(flow.leadClient.sendLatch.await(3, TimeUnit.SECONDS));
+
+        long startsBefore = agentStartCalls(flow.herdr);
+        PlacementException e = assertThrows(PlacementException.class,
+                () -> flow.workers.spawn(new SpawnRequest("terra", null, null)));
+
+        assertTrue(e.getMessage().contains("cooling off"), e.getMessage());
+        assertTrue(e.getMessage().contains("shared-openai"), e.getMessage());
+        assertEquals(startsBefore, agentStartCalls(flow.herdr),
+                "the refusal must happen before the adapter is ever called — no new agent.start");
+    }
+
+    // --- criteria 7/9: automatic placement skips every profile sharing the cooled credential -----
+
+    @Test
+    void automaticPlacementRefusesAndNamesCoolingOffWhenEveryCandidateSharesTheCredential() throws Exception {
+        Flow flow = new Flow();
+        flow.classifyBackendErrorOn(flow.s1.terminalId());
+        flow.classifyBackendErrorOn(flow.s2.terminalId());
+        assertTrue(flow.leadClient.sendLatch.await(3, TimeUnit.SECONDS));
+
+        PlacementException e = assertThrows(PlacementException.class,
+                () -> flow.workers.spawn(new SpawnRequest(null, null, null)),
+                "terra and sol both share the cooled-off credential — nothing is available");
+        assertTrue(e.getMessage().contains("cooling off"), e.getMessage());
+        assertThrows(PlacementException.class, () -> flow.workers.spawn(new SpawnRequest("sol", null, null)),
+                "sol shares terra's credential, so it must be locked out too");
+    }
+}

fleetd/src/test/java/dev/ltms/fleet/inject/CompletionResolverTest.java

@@ -768,19 +768,37 @@ class CompletionResolverTest {
     @Test
     void coverageIsOffWhenNoProfileHasAPatternConfigured() {
         assertEquals("off (no profile has an exhaustedPattern configured; profiles: [terra])",
-                CompletionResolver.coverage(Set.of("terra"), Set.of()));
+                CompletionResolver.coverage("exhaustedPattern", Set.of("terra"), Set.of()));
     }
 
     @Test
     void coverageIsFullWhenEveryProfileHasAPatternConfigured() {
         assertEquals("full (all profiles configured: [gx10, terra])",
-                CompletionResolver.coverage(Set.of("terra", "gx10"), Set.of("terra", "gx10")));
+                CompletionResolver.coverage("exhaustedPattern", Set.of("terra", "gx10"), Set.of("terra", "gx10")));
     }
 
     @Test
     void coverageIsPartialAndNamesWhichProfilesAreConfigured() {
         assertEquals("partial (configured: [terra]; not configured: [gx10])",
-                CompletionResolver.coverage(Set.of("terra", "gx10"), Set.of("terra")));
+                CompletionResolver.coverage("exhaustedPattern", Set.of("terra", "gx10"), Set.of("terra")));
+    }
+
+    /**
+     * Found live on 2026-09-03, reading a real boot log rather than a test. {@code coverage} is
+     * shared by two call sites — CB-578's {@code exhaustedPattern} line and fleetd#201 Unit 5's
+     * {@code errorPattern} line — but its "off" branch hard-coded the word {@code exhaustedPattern}.
+     * So a daemon with no {@code errorPattern} anywhere printed "no profile has an exhaustedPattern
+     * configured" directly beneath a line reporting that two profiles DO have one. Both lines were
+     * individually defensible and together they were nonsense, and the message sent an operator to
+     * set the wrong key.
+     *
+     * <p>Every earlier test here passed the exhaustion case only, so none of them could see it. This
+     * one pins that the message names the key the caller actually meant.
+     */
+    @Test
+    void coverageNamesTheConfigKeyItsCallerMeansRatherThanAlwaysSayingExhaustedPattern() {
+        assertEquals("off (no profile has an errorPattern configured; profiles: [gx10, terra])",
+                CompletionResolver.coverage("errorPattern", Set.of("terra", "gx10"), Set.of()));
     }
 
     // --- fleetd#201 Unit 1: target-keyed backend-error pattern + typed sink ----------------------

fleetd/src/test/java/dev/ltms/fleet/mcp/FleetMcpTest.java

@@ -20,6 +20,7 @@ import dev.ltms.fleet.session.MemberSession;
 import dev.ltms.fleet.session.WorktreeRequest;
 import dev.ltms.fleet.member.ClaudeCodeLauncher;
 import dev.ltms.fleet.member.CompositePeerLauncher;
+import dev.ltms.fleet.placement.BackendOutagePolicy;
 import dev.ltms.fleet.placement.BackendQuarantine;
 import dev.ltms.fleet.placement.PlacementPolicies;
 import io.modelcontextprotocol.spec.McpSchema;
@@ -500,6 +501,25 @@ class FleetMcpTest {
         assertTrue(out.contains("\"quarantinedForSeconds\":1800"), out);
     }
 
+    /** fleetd #201 Unit 5: {@code coolingOff} is a SEPARATE map from {@code quarantined}. */
+    @Test
+    void profilesReportsACoolingOffCredentialInASeparateMap() {
+        FakeHerdr h = new FakeHerdr();
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        outagePolicy.record("shared-openai", "t1", "API Error: rate limited");
+        outagePolicy.record("shared-openai", "t2", "API Error: rate limited"); // 2nd distinct target starts the incident
+        FleetMcp.OutageSource outage = new FleetMcp.OutageSource(
+                profile -> "ltms-local".equals(profile) ? "shared-openai" : null, outagePolicy);
+        McpSchema.CallToolResult res = FleetMcp.profiles(
+                workerService(h, "http://gx00.gw:8000", Set.of("gx00.gw")), FleetMcp.QuarantineSource.none(), outage);
+        assertNotEquals(Boolean.TRUE, res.isError());
+        String out = textOf(res);
+        assertTrue(out.contains("\"coolingOff\""), out);
+        assertTrue(out.contains("shared-openai"), out);
+        assertTrue(out.contains("\"coolingOffForSeconds\":60"), out);
+        assertFalse(out.contains("\"quarantined\""), "nothing is exhaustion-quarantined here: " + out);
+    }
+
     @Test
     void listReportsTrackedWorkers() {
         FakeHerdr h = new FakeHerdr();
@@ -631,6 +651,81 @@ class FleetMcpTest {
         assertEquals(2, out.split("\"free\":0", -1).length - 1, out);
     }
 
+    /** fleetd #201 Unit 5: cool-off forces {@code free:0} but never adds {@code quarantinedForSeconds}. */
+    @Test
+    void coolingOffProfileReportsZeroFreeButNeverQuarantinedForSeconds() {
+        FakeHerdr h = new FakeHerdr();
+        SessionManager sessions = new SessionManager(workerService(h, "http://gx00.gw:8000", Set.of("gx00.gw")));
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        outagePolicy.record("shared-openai", "t1", "API Error: rate limited");
+        outagePolicy.record("shared-openai", "t2", "API Error: rate limited");
+        FleetMcp.OutageSource outage = new FleetMcp.OutageSource(
+                profile -> "terra".equals(profile) ? "shared-openai" : null, outagePolicy);
+
+        String out = textOf(FleetMcp.listFleet(workerService(h, "http://gx00.gw:8000", Set.of("gx00.gw")),
+                sessions, null, new FleetMcp.CapacitySource(profile -> 0, profile -> 2,
+                () -> Set.of("terra"), () -> 0), new FleetMcp.HealthCoverageSource(() -> "off"),
+                FleetMcp.QuarantineSource.none(), outage, Map.of(), ""));
+
+        assertTrue(out.contains("\"profile\":\"terra\""), out);
+        assertTrue(out.contains("\"free\":0"), out);
+        assertTrue(out.contains("\"credentialId\":\"shared-openai\""), out);
+        assertTrue(out.contains("\"coolingOffForSeconds\":60"), out);
+        assertFalse(out.contains("quarantinedForSeconds"),
+                "exhaustion quarantine was never active — this key must not appear: " + out);
+    }
+
+    /**
+     * The two checks are independent — a profile can be BOTH exhaustion-quarantined AND cooling off
+     * for the same credential at once, and both maps/fields report it simultaneously.
+     */
+    @Test
+    void bothQuarantineAndCoolingOffCanFireForTheSameProfileAtOnce() {
+        FakeHerdr h = new FakeHerdr();
+        SessionManager sessions = new SessionManager(workerService(h, "http://gx00.gw:8000", Set.of("gx00.gw")));
+        BackendQuarantine quarantine = new BackendQuarantine(() -> 0L, TimeUnit.MINUTES.toNanos(20));
+        quarantine.quarantine("shared-openai");
+        FleetMcp.QuarantineSource quarantineSource = new FleetMcp.QuarantineSource(
+                profile -> "terra".equals(profile) ? "shared-openai" : null, quarantine);
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        outagePolicy.record("shared-openai", "t1", "API Error: rate limited");
+        outagePolicy.record("shared-openai", "t2", "API Error: rate limited");
+        FleetMcp.OutageSource outage = new FleetMcp.OutageSource(
+                profile -> "terra".equals(profile) ? "shared-openai" : null, outagePolicy);
+
+        String out = textOf(FleetMcp.listFleet(workerService(h, "http://gx00.gw:8000", Set.of("gx00.gw")),
+                sessions, null, new FleetMcp.CapacitySource(profile -> 0, profile -> 2,
+                () -> Set.of("terra"), () -> 0), new FleetMcp.HealthCoverageSource(() -> "off"),
+                quarantineSource, outage, Map.of(), ""));
+
+        assertTrue(out.contains("\"free\":0"), out);
+        assertTrue(out.contains("\"quarantinedForSeconds\":1200"), out);
+        assertTrue(out.contains("\"coolingOffForSeconds\":60"), out);
+    }
+
+    @Test
+    void listAndProfilesAgreeOnWhatIsCoolingOff() {
+        FakeHerdr h = new FakeHerdr();
+        SessionManager sessions = new SessionManager(workerService(h, "http://gx00.gw:8000", Set.of("gx00.gw")));
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        outagePolicy.record("shared-openai", "t1", "API Error: rate limited");
+        outagePolicy.record("shared-openai", "t2", "API Error: rate limited");
+        FleetMcp.OutageSource outage = new FleetMcp.OutageSource(
+                profile -> "ltms-local".equals(profile) ? "shared-openai" : null, outagePolicy);
+        var workers = workerService(h, "http://gx00.gw:8000", Set.of("gx00.gw"));
+
+        String listOut = textOf(FleetMcp.listFleet(workers, sessions, null,
+                new FleetMcp.CapacitySource(profile -> 0, profile -> 2, () -> Set.of("ltms-local"), () -> 0),
+                new FleetMcp.HealthCoverageSource(() -> "off"), FleetMcp.QuarantineSource.none(), outage,
+                Map.of(), ""));
+        String profilesOut = textOf(FleetMcp.profiles(workers, FleetMcp.QuarantineSource.none(), outage));
+
+        assertTrue(listOut.contains("\"free\":0"), listOut);
+        assertTrue(listOut.contains("\"credentialId\":\"shared-openai\""), listOut);
+        assertTrue(profilesOut.contains("\"coolingOff\""), profilesOut);
+        assertTrue(profilesOut.contains("shared-openai"), profilesOut);
+    }
+
     @Test
     void listAndProfilesAgreeOnWhatIsQuarantined() {
         FakeHerdr h = new FakeHerdr();

fleetd/src/test/java/dev/ltms/fleet/mcp/McpContractDocTest.java

@@ -0,0 +1,121 @@
+package dev.ltms.fleet.mcp;
+
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.util.LinkedHashSet;
+import java.util.Set;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+import org.junit.jupiter.api.DisplayName;
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertTrue;
+
+/**
+ * fleetd #114 (CB-609): the guard that lets {@code docs/MCP-Contract.md} name a tool at all.
+ *
+ * <p>That page was written in July 2026, before any MCP code existed, and then did not follow the
+ * code. By August it named two tools that had never been built, omitted five that shipped, had the
+ * wrong name for nearly every parameter, and still described a caller-identity rule that was a
+ * privilege bug by then. Nothing failed, because nothing checked it — and {@code CLAUDE.md} sends
+ * every session in the fleet to that page.
+ *
+ * <p>The fix was to delete the tool catalogue rather than correct it: a hand-maintained second copy
+ * of the tool surface is the defect, not the particular errors it had accumulated. What survives is
+ * the flows, which are shapes rather than names. But the flows still have to say {@code fleet_send}
+ * somewhere to be readable, and that is exactly the sentence that rots. This test is what makes it
+ * safe to write.
+ *
+ * <p><b>It checks source text, not behaviour.</b> It reads the Markdown and reads {@link FleetMcp}'s
+ * source, and it only catches a name in the doc that the server does not register. It cannot catch a
+ * flow that describes the wrong order, or a parameter name in prose — those are not name-shaped. The
+ * doc's own header carries that caveat for its readers.
+ */
+class McpContractDocTest {
+
+    /** Tests run with the module directory as cwd, so the repo-root doc is one level up. */
+    private static final Path DOC = Path.of("../docs/MCP-Contract.md");
+    private static final Path MCP_SOURCE = Path.of("src/main/java/dev/ltms/fleet/mcp/FleetMcp.java");
+
+    private static Set<String> matches(Path file, String regex) throws Exception {
+        Matcher m = Pattern.compile(regex).matcher(Files.readString(file));
+        Set<String> found = new LinkedHashSet<>();
+        while (m.find()) {
+            found.add(m.group(1));
+        }
+        return found;
+    }
+
+    /** Every {@code fleet_*} the doc mentions, in prose or in a diagram. */
+    private static Set<String> toolsNamedInTheDoc() throws Exception {
+        return matches(DOC, "(fleet_[a-z_]+)");
+    }
+
+    /** Every tool {@link FleetMcp} actually registers, read from its {@code tool("…")} calls. */
+    private static Set<String> toolsTheServerRegisters() throws Exception {
+        return matches(MCP_SOURCE, "tool\\(\"(fleet_[a-z_]+)\"");
+    }
+
+    @Test
+    @DisplayName("[SOURCE TEXT] every fleet_* tool named in MCP-Contract.md is one the server registers")
+    void theDocNamesNoToolThatDoesNotExist() throws Exception {
+        Set<String> registered = toolsTheServerRegisters();
+        Set<String> named = toolsNamedInTheDoc();
+
+        Set<String> unknown = new LinkedHashSet<>(named);
+        unknown.removeAll(registered);
+
+        assertTrue(unknown.isEmpty(),
+                "docs/MCP-Contract.md names " + unknown + ", which FleetMcp does not register. "
+                        + "Checked " + named.size() + " name(s) in the doc against " + registered.size()
+                        + " registered tool(s): " + registered + ". This is the fleetd #114 defect "
+                        + "recurring — the doc named fleet_read and fleet_cancel for weeks after the "
+                        + "code shipped without them. Either fix the name or drop it from the page; do "
+                        + "NOT weaken this test.");
+    }
+
+    /**
+     * The denominator guard. The check above passes trivially if the doc stops naming any tool at
+     * all — an empty set is a subset of everything. A checker that can silently check nothing is the
+     * fleetd #113 shape, so this pins that the doc really is still describing the flows, and that
+     * the registration scrape really did find the server's tools.
+     */
+    @Test
+    @DisplayName("[SOURCE TEXT] the doc/server name check is not vacuous — both sides found names")
+    void theCheckActuallyHasSomethingToCheck() throws Exception {
+        Set<String> registered = toolsTheServerRegisters();
+        Set<String> named = toolsNamedInTheDoc();
+
+        assertTrue(registered.size() >= 10,
+                "scraped only " + registered.size() + " tool registrations from FleetMcp (" + registered
+                        + "); the server registers eleven, so the tool(\"…\") scrape has stopped matching "
+                        + "and the check above is now vacuous");
+        assertTrue(named.size() >= 4,
+                "docs/MCP-Contract.md names only " + named.size() + " fleet_* tool(s) (" + named + "). "
+                        + "The flows describe delegation, clarification, detached delivery and the "
+                        + "turn-done fallback, so it should name several. Too few means the page has been "
+                        + "gutted and this test is guarding nothing.");
+    }
+
+    /**
+     * fleetd #114's actual lesson. The catalogue was deleted on purpose; a well-meaning "let me just
+     * document the tools here" restores the exact second copy that drifted for a month.
+     */
+    @Test
+    @DisplayName("[SOURCE TEXT] MCP-Contract.md still says it is not the tool reference")
+    void theDocStillDisclaimsBeingTheToolReference() throws Exception {
+        String doc = Files.readString(DOC);
+        assertTrue(doc.contains("**What this page is NOT: a tool reference.**"),
+                "docs/MCP-Contract.md must keep saying it is not the tool reference. That sentence is "
+                        + "the fix for fleetd #114: the page carried a hand-maintained tool catalogue that "
+                        + "drifted from the code for a month while CLAUDE.md pointed every session at it.");
+        assertEquals(0, countTables(doc.substring(0, doc.indexOf("## 1. Rendezvous flows"))),
+                "the header of docs/MCP-Contract.md must not grow a tool/parameter table — that is the "
+                        + "second copy fleetd #114 deleted");
+    }
+
+    private static int countTables(String markdown) {
+        return (int) markdown.lines().filter(l -> l.strip().startsWith("|")).count();
+    }
+}

fleetd/src/test/java/dev/ltms/fleet/member/ClaudeCodeLauncherTest.java

@@ -4,6 +4,9 @@ import ch.qos.logback.classic.Level;
 import ch.qos.logback.classic.Logger;
 import ch.qos.logback.classic.spi.ILoggingEvent;
 import ch.qos.logback.core.read.ListAppender;
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import com.fasterxml.jackson.databind.node.ObjectNode;
 import dev.ltms.fleet.config.FleetConfig;
 import dev.ltms.fleet.guard.GuardException;
 import dev.ltms.fleet.guard.SubscriptionGuard;
@@ -23,11 +26,16 @@ import org.slf4j.LoggerFactory;
 import java.io.IOException;
 import java.nio.file.Files;
 import java.nio.file.Path;
+import java.nio.file.attribute.PosixFileAttributeView;
 import java.nio.file.attribute.PosixFilePermissions;
+import java.util.HashSet;
 import java.util.List;
 import java.util.Map;
 import java.util.Set;
 import java.util.UUID;
+import java.util.concurrent.CountDownLatch;
+import java.util.concurrent.TimeUnit;
+import java.util.concurrent.atomic.AtomicBoolean;
 import java.util.concurrent.atomic.AtomicReference;
 import java.util.function.Function;
 import java.util.function.Supplier;
@@ -2098,4 +2106,437 @@ class ClaudeCodeLauncherTest {
         assertTrue(charterFile.getFileName().toString().startsWith("fleetd-role-charter-"),
                 "same file-naming scheme as before this fix (no wrapping directory): " + charterFile);
     }
+
+    // --- fleetd #149: workspace-trust dialog seed ------------------------------------------------
+    //
+    // Claude Code asks an interactive, un-timed "Is this a project you created or one you trust?"
+    // the first time it starts in a directory it has not seen. Every worktree: true spawn lands in
+    // a brand-new directory, so without a seed the member sits on that dialog forever — herdr still
+    // reports it healthy (agent_status: blocked, interactive_ready: true) — and never mounts the
+    // bridge MCP or calls fleet_reply. These tests start the REAL launcher (only the herdr transport
+    // is faked) so the seed is proven to run inside buildLaunch, before agent.start (the
+    // process-starting call) ever fires — a test that only checked the JSON writer in isolation
+    // would prove nothing about whether the launcher actually calls it at the right time.
+    //
+    // INCIDENT: the seed originally ran on ANY non-blank cwd. Running this file's own test suite —
+    // most of whose fixtures spawn with no cwd/configDir set, so both fall back to the real
+    // user.dir / ~/.claude.json — corrupted the operator's actual ~/.claude.json (it shrank from
+    // ~72 KB to a single seeded entry) the first time a manual mutation run made the write
+    // non-additive. The fix restricts the seed to isProvisionedWorktree(cwd) — a real .git FILE
+    // (not directory) — exactly the writeIdeOverlay gate already used for the same category of
+    // risk. Every test below marks its own worktree fixture with that .git file, and
+    // seedTrustDialogNeverWritesWhenCwdIsNotAProvisionedWorktree is the regression test for the
+    // incident itself.
+
+    private FleetConfig.Profile trustProfile(String configDir, String cwd) {
+        return new FleetConfig.Profile(
+                "ltms-local", "http://gx00.gw:8000", "coder", configDir, "FLEETD_WORKER_TOKEN",
+                List.of("claude"), "tab", "fleetd-workers", "w #{n}", "http://127.0.0.1:8765/mcp",
+                cwd, null);
+    }
+
+    /**
+     * Give {@code dir} the exact signature {@link ClaudeCodeLauncher#isProvisionedWorktree} (and
+     * {@code writeIdeOverlay} before it) checks for: a {@code .git} REGULAR FILE, never a
+     * directory. The content is never parsed by the trust seed, so any {@code gitdir:} pointer is
+     * fine.
+     */
+    private static void markAsProvisionedWorktree(Path dir) throws IOException {
+        Files.writeString(dir.resolve(".git"), "gitdir: /tmp/not-a-real-gitdir");
+    }
+
+    @Test
+    void seedsWorkspaceTrustForTheCwdBeforeTheProcessStarts(
+            @TempDir Path configDir, @TempDir Path worktree) throws Exception {
+        markAsProvisionedWorktree(worktree);
+        FakeHerdr herdr = new FakeHerdr();
+        Path claudeJson = configDir.resolve(".claude.json");
+        AtomicReference<Boolean> seededBeforeStart = new AtomicReference<>(false);
+        herdr.onAgentStart(() -> {
+            try {
+                if (!Files.exists(claudeJson)) {
+                    return;
+                }
+                JsonNode root = new ObjectMapper().readTree(claudeJson.toFile());
+                JsonNode project = root.path("projects").path(worktree.toString());
+                seededBeforeStart.set(project.path("hasTrustDialogAccepted").asBoolean(false));
+            } catch (IOException e) {
+                seededBeforeStart.set(false);
+            }
+        });
+
+        FleetConfig.Profile cfg = trustProfile(configDir.toString(), worktree.toString());
+        new ClaudeCodeLauncher(new AgentControl(herdr), new WorkspaceControl(herdr),
+                new SubscriptionGuard(Set.of("gx00.gw")), Map.of(cfg.profile(), cfg), cfg.profile(), _ -> null)
+                .spawn();
+
+        assertTrue(herdr.called("agent.start"), "the hook must actually have fired during the spawn");
+        assertEquals(Boolean.TRUE, seededBeforeStart.get(),
+                "the trust entry for the cwd must already exist at the instant agent.start (the "
+                        + "process-starting herdr call) fires — not merely once spawn() returns");
+    }
+
+    @Test
+    void seedTrustDialogWritesOnlyTheTrustFlagAndNotTheOnboardingKey(
+            @TempDir Path configDir, @TempDir Path worktree) throws Exception {
+        markAsProvisionedWorktree(worktree);
+        FakeHerdr herdr = new FakeHerdr();
+        FleetConfig.Profile cfg = trustProfile(configDir.toString(), worktree.toString());
+        new ClaudeCodeLauncher(new AgentControl(herdr), new WorkspaceControl(herdr),
+                new SubscriptionGuard(Set.of("gx00.gw")), Map.of(cfg.profile(), cfg), cfg.profile(), _ -> null)
+                .spawn();
+
+        Path claudeJson = configDir.resolve(".claude.json");
+        assertTrue(Files.exists(claudeJson), "seeded into <configDir>/.claude.json");
+        JsonNode project = new ObjectMapper().readTree(claudeJson.toFile())
+                .path("projects").path(worktree.toString());
+        assertTrue(project.path("hasTrustDialogAccepted").asBoolean(false));
+        // fleetd #247: the onboarding key must NOT be written. Claude Code strips it on every
+        // save (measured: 0 of 28 live entries had it, including its own), so writing it only
+        // adds a contested key to a file two processes share. This assertion is the guard that
+        // stops it coming back as a plausible-looking "completeness" fix.
+        assertFalse(project.has("hasCompletedProjectOnboarding"),
+                "hasCompletedProjectOnboarding must not be written — Claude Code drops it on "
+                        + "every save, and the member reaches idle on hasTrustDialogAccepted alone");
+    }
+
+    /**
+     * Criterion 3: seeding is additive. An existing {@code .claude.json} carries the operator's own
+     * project history and unrelated top-level settings — the seed must change only
+     * {@code projects.<cwd>} for THIS cwd and leave everything else, including a different project's
+     * own unrelated data, exactly as it was.
+     */
+    @Test
+    void seedTrustDialogIsAdditiveAndPreservesUnknownKeysAndOtherProjects(
+            @TempDir Path configDir, @TempDir Path worktree) throws Exception {
+        markAsProvisionedWorktree(worktree);
+        Path claudeJson = configDir.resolve(".claude.json");
+        Files.writeString(claudeJson, """
+                {
+                  "numStartups": 42,
+                  "oauthAccount": {"emailAddress": "operator@example.com"},
+                  "projects": {
+                    "/some/other/project": {
+                      "hasTrustDialogAccepted": true,
+                      "mcpServers": {"foo": {"type": "stdio", "command": "foo-mcp"}}
+                    }
+                  }
+                }
+                """);
+
+        FakeHerdr herdr = new FakeHerdr();
+        FleetConfig.Profile cfg = trustProfile(configDir.toString(), worktree.toString());
+        new ClaudeCodeLauncher(new AgentControl(herdr), new WorkspaceControl(herdr),
+                new SubscriptionGuard(Set.of("gx00.gw")), Map.of(cfg.profile(), cfg), cfg.profile(), _ -> null)
+                .spawn();
+
+        JsonNode root = new ObjectMapper().readTree(claudeJson.toFile());
+        assertEquals(42, root.path("numStartups").asInt(), "unrelated top-level key survives untouched");
+        assertEquals("operator@example.com", root.path("oauthAccount").path("emailAddress").asText(),
+                "an unrelated nested top-level key survives untouched");
+
+        JsonNode other = root.path("projects").path("/some/other/project");
+        assertTrue(other.path("hasTrustDialogAccepted").asBoolean(false),
+                "a different project's own trust entry survives");
+        assertEquals("foo-mcp", other.path("mcpServers").path("foo").path("command").asText(),
+                "a different project's own unrelated nested data survives");
+
+        JsonNode mine = root.path("projects").path(worktree.toString());
+        assertTrue(mine.path("hasTrustDialogAccepted").asBoolean(false));
+    }
+
+    /**
+     * Criterion 2: where the profile sets no {@code configDir}, the seed goes to the default
+     * {@code ~/.claude.json}. {@code user.home} is redirected to a {@code @TempDir} for the
+     * duration of this test and restored in a {@code finally} — the real operator {@code
+     * ~/.claude.json} must never be touched by a test.
+     */
+    @Test
+    void seedTrustDialogTargetsDefaultClaudeJsonWhenConfigDirIsUnset(
+            @TempDir Path fakeHome, @TempDir Path worktree) throws Exception {
+        markAsProvisionedWorktree(worktree);
+        String originalHome = System.getProperty("user.home");
+        System.setProperty("user.home", fakeHome.toString());
+        try {
+            FakeHerdr herdr = new FakeHerdr();
+            FleetConfig.Profile cfg = trustProfile(null, worktree.toString());
+            new ClaudeCodeLauncher(new AgentControl(herdr), new WorkspaceControl(herdr),
+                    new SubscriptionGuard(Set.of("gx00.gw")), Map.of(cfg.profile(), cfg), cfg.profile(), _ -> null)
+                    .spawn();
+
+            Path claudeJson = fakeHome.resolve(".claude.json");
+            assertTrue(Files.exists(claudeJson),
+                    "no configDir set — the default target is ~/.claude.json, here the redirected fake home");
+            JsonNode project = new ObjectMapper().readTree(claudeJson.toFile())
+                    .path("projects").path(worktree.toString());
+            assertTrue(project.path("hasTrustDialogAccepted").asBoolean(false));
+        } finally {
+            System.setProperty("user.home", originalHome);
+        }
+    }
+
+    /**
+     * Regression test for the fleetd #149 incident itself: a cwd that is NOT a fleetd-provisioned
+     * worktree (no {@code .git} FILE — the exact shape a real checkout, or an un-configured
+     * fallback cwd, has) must never be written to, however {@code configDir} is set. This is the
+     * fix for the exact defect that corrupted the operator's real {@code ~/.claude.json}.
+     */
+    @Test
+    void seedTrustDialogNeverWritesWhenCwdIsNotAProvisionedWorktree(
+            @TempDir Path configDir, @TempDir Path plainCwd) {
+        // plainCwd deliberately carries NO .git file — the same shape a real checkout's cwd
+        // fallback has.
+        FakeHerdr herdr = new FakeHerdr();
+        FleetConfig.Profile cfg = trustProfile(configDir.toString(), plainCwd.toString());
+        new ClaudeCodeLauncher(new AgentControl(herdr), new WorkspaceControl(herdr),
+                new SubscriptionGuard(Set.of("gx00.gw")), Map.of(cfg.profile(), cfg), cfg.profile(), _ -> null)
+                .spawn();
+
+        assertFalse(Files.exists(configDir.resolve(".claude.json")),
+                "a non-worktree cwd must never get a .claude.json written for it — this is the fix "
+                        + "for the incident where writing unconditionally corrupted the operator's own "
+                        + "real ~/.claude.json via this file's own no-cwd/no-configDir test fixtures");
+    }
+
+    /**
+     * Same regression, for the default (no {@code configDir}) path — the exact combination (no
+     * {@code configDir}, no worktree-shaped {@code cwd}) that hit the operator's real
+     * {@code ~/.claude.json} during the incident. {@code user.home} is still redirected to a
+     * {@code @TempDir} out of caution, so even a reintroduced bug here cannot touch the real file.
+     */
+    @Test
+    void seedTrustDialogNeverWritesToDefaultHomeWhenCwdIsNotAProvisionedWorktree(
+            @TempDir Path fakeHome, @TempDir Path plainCwd) {
+        String originalHome = System.getProperty("user.home");
+        System.setProperty("user.home", fakeHome.toString());
+        try {
+            FakeHerdr herdr = new FakeHerdr();
+            FleetConfig.Profile cfg = trustProfile(null, plainCwd.toString());
+            new ClaudeCodeLauncher(new AgentControl(herdr), new WorkspaceControl(herdr),
+                    new SubscriptionGuard(Set.of("gx00.gw")), Map.of(cfg.profile(), cfg), cfg.profile(), _ -> null)
+                    .spawn();
+
+            assertFalse(Files.exists(fakeHome.resolve(".claude.json")),
+                    "the exact incident combination — no configDir, non-worktree cwd — must never "
+                            + "write, even to the (redirected) default ~/.claude.json");
+        } finally {
+            System.setProperty("user.home", originalHome);
+        }
+    }
+
+    // --- fleetd #149 review round 2: the write must be atomic and lock-protected -----------------
+    //
+    // The first round fixed WHICH cwd this can ever target. This round fixes HOW the target is
+    // written: .claude.json is large (tens of KB, dozens of projects on a real host), Claude Code
+    // itself rewrites it while running, and this daemon spawns several members in parallel as a
+    // matter of routine. The pre-fix Files.writeString(target, content) truncates target in place
+    // before writing the replacement bytes — a crash mid-write, or another writer's read landing in
+    // that window, loses data. Two more failure modes follow directly: (a) a crash/kill mid-write
+    // leaves target truncated, and (b) two concurrent spawns racing a naive read-modify-write let
+    // the second writer's write silently discard the first spawn's entry. The fix is
+    // ClaudeCodeLauncher.writeAtomically (sibling temp file + ATOMIC_MOVE, package-visible for the
+    // test below that proves it directly) plus TRUST_JSON_LOCK (a process-wide lock serialising
+    // every seedTrustDialog call this launcher itself makes).
+
+    /**
+     * Requested test 1: an existing, large-ish (not just a two-key fixture) .claude.json must never
+     * collapse. Asserts on the restored KEY SET — not merely that the result still parses as JSON,
+     * which the incident's 178-byte file also did.
+     */
+    @Test
+    void seedTrustDialogPreservesALargeExistingFileWithoutCollapsing(
+            @TempDir Path configDir, @TempDir Path worktree) throws Exception {
+        markAsProvisionedWorktree(worktree);
+        Path claudeJson = configDir.resolve(".claude.json");
+
+        ObjectMapper mapper = new ObjectMapper();
+        ObjectNode root = mapper.createObjectNode();
+        root.put("numStartups", 4200);
+        root.put("firstStartTime", "2025-01-01T00:00:00.000Z");
+        root.putObject("oauthAccount").put("emailAddress", "operator@example.com");
+        Set<String> otherProjectPaths = new HashSet<>();
+        ObjectNode projects = root.putObject("projects");
+        for (int i = 0; i < 30; i++) {
+            String path = "/Users/operator/code/project-" + i;
+            otherProjectPaths.add(path);
+            ObjectNode project = projects.putObject(path);
+            project.put("hasTrustDialogAccepted", true);
+            project.putObject("mcpServers").putObject("server-" + i).put("command", "server-" + i + "-mcp");
+        }
+        String before = mapper.writerWithDefaultPrettyPrinter().writeValueAsString(root);
+        Files.writeString(claudeJson, before);
+        long sizeBefore = Files.size(claudeJson);
+        assertTrue(sizeBefore > 4096,
+                "fixture must actually be large-ish to be a meaningful proof: " + sizeBefore + " bytes");
+
+        FakeHerdr herdr = new FakeHerdr();
+        FleetConfig.Profile cfg = trustProfile(configDir.toString(), worktree.toString());
+        new ClaudeCodeLauncher(new AgentControl(herdr), new WorkspaceControl(herdr),
+                new SubscriptionGuard(Set.of("gx00.gw")), Map.of(cfg.profile(), cfg), cfg.profile(), _ -> null)
+                .spawn();
+
+        long sizeAfter = Files.size(claudeJson);
+        assertTrue(sizeAfter >= sizeBefore,
+                "the file must never collapse below its pre-seed size — before=" + sizeBefore
+                        + " after=" + sizeAfter + " bytes (the incident shrank ~72 KB to 178 bytes)");
+
+        JsonNode after = mapper.readTree(claudeJson.toFile());
+        assertEquals(4200, after.path("numStartups").asInt(), "unrelated top-level key survives");
+        assertEquals("operator@example.com", after.path("oauthAccount").path("emailAddress").asText());
+        JsonNode afterProjects = after.path("projects");
+        for (String path : otherProjectPaths) {
+            assertTrue(afterProjects.path(path).path("hasTrustDialogAccepted").asBoolean(false),
+                    "pre-existing project entry " + path + " must survive");
+        }
+        assertEquals(otherProjectPaths.size() + 1, afterProjects.size(),
+                "exactly one NEW project entry (this worktree's) must be added, none dropped");
+        assertTrue(afterProjects.path(worktree.toString()).path("hasTrustDialogAccepted").asBoolean(false));
+    }
+
+    /**
+     * Requested test 2: two concurrent spawns for DIFFERENT cwd values sharing one configDir must
+     * both end up present in the final file — a naive concurrent read-modify-write would let the
+     * second writer's read (taken before the first writer's write lands) silently discard the
+     * first. A {@link CountDownLatch} — not a sleep — lines both threads up at the starting line so
+     * this does not depend on scheduling luck to be meaningful.
+     *
+     * <p>This is a test of {@code TRUST_JSON_LOCK}, not of {@code writeAtomically}: the lock fully
+     * serialises every {@code seedTrustDialog} call this launcher itself makes, so this test would
+     * pass even without atomicity. See {@link #writeAtomicallyNeverExposesATornFileToAConcurrentReader}
+     * for the test that exercises atomicity specifically.
+     */
+    @Test
+    void concurrentSeedsForDifferentCwdsBothSurvive(
+            @TempDir Path configDir, @TempDir Path worktreeA, @TempDir Path worktreeB) throws Exception {
+        markAsProvisionedWorktree(worktreeA);
+        markAsProvisionedWorktree(worktreeB);
+
+        CountDownLatch ready = new CountDownLatch(2);
+        CountDownLatch go = new CountDownLatch(1);
+        AtomicReference<Exception> failureA = new AtomicReference<>();
+        AtomicReference<Exception> failureB = new AtomicReference<>();
+
+        Thread ta = new Thread(spawnTask(configDir, worktreeA, ready, go, failureA), "spawn-a");
+        Thread tb = new Thread(spawnTask(configDir, worktreeB, ready, go, failureB), "spawn-b");
+        ta.start();
+        tb.start();
+
+        assertTrue(ready.await(5, TimeUnit.SECONDS), "both threads must reach the starting line");
+        go.countDown();
+        ta.join(5000);
+        tb.join(5000);
+        assertFalse(ta.isAlive(), "spawn A must finish within the timeout");
+        assertFalse(tb.isAlive(), "spawn B must finish within the timeout");
+        assertNull(failureA.get(), "spawn A must not throw: " + failureA.get());
+        assertNull(failureB.get(), "spawn B must not throw: " + failureB.get());
+
+        JsonNode root = new ObjectMapper().readTree(configDir.resolve(".claude.json").toFile());
+        assertTrue(root.path("projects").path(worktreeA.toString())
+                        .path("hasTrustDialogAccepted").asBoolean(false),
+                "worktree A's entry must survive the race");
+        assertTrue(root.path("projects").path(worktreeB.toString())
+                        .path("hasTrustDialogAccepted").asBoolean(false),
+                "worktree B's entry must survive the race");
+    }
+
+    private Runnable spawnTask(Path configDir, Path worktree, CountDownLatch ready, CountDownLatch go,
+                                AtomicReference<Exception> failure) {
+        return () -> {
+            try {
+                FakeHerdr herdr = new FakeHerdr();
+                FleetConfig.Profile cfg = trustProfile(configDir.toString(), worktree.toString());
+                ClaudeCodeLauncher launcher = new ClaudeCodeLauncher(new AgentControl(herdr),
+                        new WorkspaceControl(herdr), new SubscriptionGuard(Set.of("gx00.gw")),
+                        Map.of(cfg.profile(), cfg), cfg.profile(), _ -> null);
+                ready.countDown();
+                go.await();
+                launcher.spawn();
+            } catch (Exception e) {
+                failure.set(e);
+            }
+        };
+    }
+
+    /**
+     * Requested test 3: the atomic write must preserve {@code .claude.json}'s existing {@code 0600}
+     * permissions, not silently widen them via a fresh temp file's own defaults landing on top of a
+     * file that had different (e.g. group-readable) permissions. Skips rather than fails on a
+     * filesystem with no POSIX permissions (e.g. Windows) — the same {@code assumeTrue} pattern this
+     * file already uses for {@code memberHerdrSocketWithWorktreeRootAndGroupPutsCharterUnderWorktreeRootAndSharesIt}.
+     */
+    @Test
+    void seedTrustDialogPreservesExisting0600Permissions(
+            @TempDir Path configDir, @TempDir Path worktree) throws Exception {
+        markAsProvisionedWorktree(worktree);
+        Path claudeJson = configDir.resolve(".claude.json");
+        Files.writeString(claudeJson, "{}");
+        assumeTrue(Files.getFileAttributeView(claudeJson, PosixFileAttributeView.class) != null,
+                "no POSIX permissions on this filesystem — skipping rather than failing");
+        Files.setPosixFilePermissions(claudeJson, PosixFilePermissions.fromString("rw-------"));
+
+        FakeHerdr herdr = new FakeHerdr();
+        FleetConfig.Profile cfg = trustProfile(configDir.toString(), worktree.toString());
+        new ClaudeCodeLauncher(new AgentControl(herdr), new WorkspaceControl(herdr),
+                new SubscriptionGuard(Set.of("gx00.gw")), Map.of(cfg.profile(), cfg), cfg.profile(), _ -> null)
+                .spawn();
+
+        assertEquals("rw-------", PosixFilePermissions.toString(Files.getPosixFilePermissions(claudeJson)),
+                "the atomic write must preserve .claude.json's existing 0600 permissions, not widen "
+                        + "them via a fresh temp file's own defaults");
+    }
+
+    /**
+     * Direct proof that {@link ClaudeCodeLauncher#writeAtomically} — not {@code TRUST_JSON_LOCK} —
+     * is what keeps a concurrent reader of {@code target} from ever observing a truncated or
+     * partially-written file. Calls {@code writeAtomically} directly (package-visible for exactly
+     * this test) rather than going through {@code seedTrustDialog}/{@code spawn()}, because {@link
+     * #concurrentSeedsForDifferentCwdsBothSurvive} above is protected by the lock and would pass
+     * even without atomicity — it proves the lock, not the atomic move. This test proves the atomic
+     * move specifically: a reader racing a writer OUTSIDE that lock (a second daemon process, or the
+     * operator's own live Claude Code — exactly what the lock cannot reach) must still never see a
+     * torn file.
+     *
+     * <p>The new content is made large (tens of MB) so a naive truncate-then-write has a real,
+     * non-instantaneous window for the busy-poll reader thread to land in — this is inherently a
+     * race, not a guaranteed-deterministic assertion, but it uses no {@code Thread.sleep} and
+     * reliably reproduced the torn read when run against the pre-fix
+     * {@code Files.writeString(target, content)} implementation (see the PR's mutation table).
+     */
+    @Test
+    void writeAtomicallyNeverExposesATornFileToAConcurrentReader(@TempDir Path dir) throws Exception {
+        Path target = dir.resolve(".claude.json");
+        String oldContent = "{\"marker\":\"OLD\"}";
+        Files.writeString(target, oldContent);
+        String newContent = "{\"marker\":\"NEW\",\"pad\":\"" + "x".repeat(20_000_000) + "\"}";
+
+        AtomicReference<String> tornRead = new AtomicReference<>();
+        AtomicBoolean stop = new AtomicBoolean(false);
+        Thread reader = new Thread(() -> {
+            while (!stop.get()) {
+                try {
+                    String seen = Files.readString(target);
+                    if (!seen.equals(oldContent) && !seen.equals(newContent)) {
+                        tornRead.compareAndSet(null, "torn read of length " + seen.length() + ": "
+                                + seen.substring(0, Math.min(seen.length(), 80)));
+                        stop.set(true);
+                    }
+                } catch (IOException ignored) {
+                    // ATOMIC_MOVE guarantees the path always resolves to old-or-new content once
+                    // readable at all — a transient "briefly missing during the rename" is fine to
+                    // ignore and keep sampling.
+                }
+            }
+        });
+        reader.start();
+        try {
+            ClaudeCodeLauncher.writeAtomically(target, newContent);
+        } finally {
+            stop.set(true);
+            reader.join(5000);
+        }
+
+        assertNull(tornRead.get(), "a concurrent reader must never observe a partially-written file: "
+                + tornRead.get());
+        assertEquals(newContent, Files.readString(target), "the final content must be the new content");
+    }
 }

fleetd/src/test/java/dev/ltms/fleet/member/CompositePeerLauncherTest.java

@@ -17,6 +17,7 @@ import dev.ltms.fleet.peer.PeerHandle;
 import dev.ltms.fleet.peer.PeerLauncher;
 import dev.ltms.fleet.peer.PeerUnreachableException;
 import dev.ltms.fleet.peer.SpawnRequest;
+import dev.ltms.fleet.placement.BackendOutagePolicy;
 import dev.ltms.fleet.placement.BackendQuarantine;
 import dev.ltms.fleet.placement.PlacementException;
 import dev.ltms.fleet.placement.PlacementPolicies;
@@ -911,4 +912,185 @@ class CompositePeerLauncherTest {
         assertDoesNotThrow(() -> composite.spawn(new SpawnRequest("claude", null, null)));
         assertDoesNotThrow(() -> composite.spawn(new SpawnRequest("gemini", null, null)));
     }
+
+    // ── fleetd #201 Unit 5: repeated backend errors cool a credential off — a SEPARATE, shorter- ──
+    // ── lived mechanism from CB-578 stage B quarantine above, never merged with it ───────────────
+
+    @Test
+    void explicitSpawnOntoACoolingOffProfileIsRefusedNamingTheCredentialAndRemainingTime() {
+        FakeHerdr herdr = new FakeHerdr();
+        Map<String, FleetConfig.Profile> profiles = ordered(
+                "sol", stubWorker("sol", "shared-openai"),
+                "terra", stubWorker("terra", "shared-openai"));
+        StubLauncher adapter = new StubLauncher("claude", herdr, profiles, "sol", Set.of());
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        // Two DISTINCT targets on the same credential — evidenceCount() counts distinct targets,
+        // never raw events, so one target repeating a classified error can never mint an incident.
+        outagePolicy.record("shared-openai", "term-1", "API Error: 500");
+        assertTrue(outagePolicy.record("shared-openai", "term-2", "API Error: 500").isPresent(),
+                "the second distinct target crosses the threshold and starts the cool-off");
+        CompositePeerLauncher composite = new CompositePeerLauncher(List.of(adapter), "sol", profiles,
+                PlacementPolicies.fixed(), _ -> 0, null, BackendQuarantine.none(), outagePolicy);
+
+        PlacementException e = assertThrows(PlacementException.class,
+                () -> composite.spawn(new SpawnRequest("sol", null, null)));
+        assertTrue(e.getMessage().contains("sol"), "message names the profile: " + e.getMessage());
+        assertTrue(e.getMessage().contains("shared-openai"), "message names the credential: " + e.getMessage());
+        assertTrue(e.getMessage().contains("cooling off"), "message says cooling off, not exhausted: " + e.getMessage());
+        assertTrue(e.getMessage().contains("60"), "message names roughly when it lifts: " + e.getMessage());
+        assertEquals(0, adapter.spawnCount("sol"), "the cooling-off profile is never delegated to");
+    }
+
+    /**
+     * A single error from one target must never cool a credential off — the threshold is on
+     * distinct targets. This proves the spawn gate really reads {@link BackendOutagePolicy}'s
+     * result rather than reacting to any recorded event.
+     */
+    @Test
+    void explicitSpawnIsNotRefusedAfterOnlyOneBackendErrorOnOneTarget() {
+        FakeHerdr herdr = new FakeHerdr();
+        Map<String, FleetConfig.Profile> profiles = ordered(
+                "sol", stubWorker("sol", "shared-openai"),
+                "terra", stubWorker("terra", "shared-openai"));
+        StubLauncher adapter = new StubLauncher("claude", herdr, profiles, "sol", Set.of());
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        assertTrue(outagePolicy.record("shared-openai", "term-1", "API Error: 500").isEmpty(),
+                "one distinct target is below THRESHOLD=2 — no incident yet");
+        CompositePeerLauncher composite = new CompositePeerLauncher(List.of(adapter), "sol", profiles,
+                PlacementPolicies.fixed(), _ -> 0, null, BackendQuarantine.none(), outagePolicy);
+
+        assertDoesNotThrow(() -> composite.spawn(new SpawnRequest("sol", null, null)),
+                "one target's error never cools the credential off");
+        assertEquals(1, adapter.spawnCount("sol"));
+    }
+
+    /**
+     * sol and terra share one OpenAI credential; cooling off because of an outage classified via
+     * ONE of them must lock out the other too, exactly like CB-578 stage B quarantine does.
+     */
+    @Test
+    void twoProfilesSharingACredentialAreBothCoolingOffByOneOutageIncident() {
+        FakeHerdr herdr = new FakeHerdr();
+        Map<String, FleetConfig.Profile> profiles = ordered(
+                "sol", stubWorker("sol", "shared-openai"),
+                "terra", stubWorker("terra", "shared-openai"));
+        StubLauncher adapter = new StubLauncher("claude", herdr, profiles, "sol", Set.of());
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        outagePolicy.record("shared-openai", "term-1", "API Error: 500");
+        outagePolicy.record("shared-openai", "term-2", "API Error: 500");
+        CompositePeerLauncher composite = new CompositePeerLauncher(List.of(adapter), "sol", profiles,
+                PlacementPolicies.fixed(), _ -> 0, null, BackendQuarantine.none(), outagePolicy);
+
+        assertThrows(PlacementException.class, () -> composite.spawn(new SpawnRequest("sol", null, null)));
+        assertThrows(PlacementException.class, () -> composite.spawn(new SpawnRequest("terra", null, null)),
+                "terra shares sol's credential, so it must be locked out too");
+        assertEquals(0, adapter.spawnCount("sol"));
+        assertEquals(0, adapter.spawnCount("terra"));
+    }
+
+    @Test
+    void placementSkipsACoolingOffProfileAndRoutesToAnotherOne() {
+        FakeHerdr herdr = new FakeHerdr();
+        Map<String, FleetConfig.Profile> profiles = ordered(
+                "sol", stubWorker("sol", "shared-openai"),
+                "b", stubWorker("b"));
+        StubLauncher adapter = new StubLauncher("claude", herdr, profiles, "sol", Set.of());
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        outagePolicy.record("shared-openai", "term-1", "API Error: 500");
+        outagePolicy.record("shared-openai", "term-2", "API Error: 500");
+        CompositePeerLauncher composite = new CompositePeerLauncher(List.of(adapter), "sol", profiles,
+                PlacementPolicies.weighted(), _ -> 0, null, BackendQuarantine.none(), outagePolicy);
+
+        PeerHandle h = composite.spawn(new SpawnRequest(null, null, null));
+        assertEquals("b", h.profile(), "sol is cooling off, so an unqualified spawn must land on b");
+        assertEquals(0, adapter.spawnCount("sol"));
+    }
+
+    @Test
+    void automaticPlacementNamesCoolingOffWhenEveryCandidateIsCoolingOff() {
+        FakeHerdr herdr = new FakeHerdr();
+        Map<String, FleetConfig.Profile> profiles = ordered(
+                "sol", stubWorker("sol", "shared-openai"),
+                "terra", stubWorker("terra", "shared-openai"));
+        StubLauncher adapter = new StubLauncher("claude", herdr, profiles, "sol", Set.of());
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        outagePolicy.record("shared-openai", "term-1", "API Error: 500");
+        outagePolicy.record("shared-openai", "term-2", "API Error: 500");
+        CompositePeerLauncher composite = new CompositePeerLauncher(List.of(adapter), "sol", profiles,
+                PlacementPolicies.weighted(), _ -> 0, null, BackendQuarantine.none(), outagePolicy);
+
+        // PlacementPolicy.select throws uncaught out of spawn() when available() is empty — never
+        // wrapped in PeerUnreachableException, which is only for a delegate that actually refused.
+        PlacementException e = assertThrows(PlacementException.class,
+                () -> composite.spawn(new SpawnRequest(null, null, null)),
+                "both candidates share the cooling-off credential — nothing is available");
+        assertTrue(e.getMessage().contains("cooling off"),
+                "message says cooling off, not exhausted, since nothing here is quarantined: " + e.getMessage());
+    }
+
+    @Test
+    void aCoolOffLiftsOnTheInjectedClockAndTheProfileBecomesSpawnableAgain() {
+        FakeHerdr herdr = new FakeHerdr();
+        Map<String, FleetConfig.Profile> profiles = ordered(
+                "sol", stubWorker("sol", "shared-openai"),
+                "b", stubWorker("b"));
+        StubLauncher adapter = new StubLauncher("claude", herdr, profiles, "sol", Set.of());
+        AtomicLong nowNanos = new AtomicLong(0L);
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(nowNanos::get);
+        outagePolicy.record("shared-openai", "term-1", "API Error: 500");
+        outagePolicy.record("shared-openai", "term-2", "API Error: 500");
+        CompositePeerLauncher composite = new CompositePeerLauncher(List.of(adapter), "sol", profiles,
+                PlacementPolicies.fixed(), _ -> 0, null, BackendQuarantine.none(), outagePolicy);
+
+        assertThrows(PlacementException.class, () -> composite.spawn(new SpawnRequest("sol", null, null)),
+                "still inside the 60s cool-off");
+
+        nowNanos.set(TimeUnit.SECONDS.toNanos(61));
+
+        PeerHandle h = composite.spawn(new SpawnRequest("sol", null, null));
+        assertEquals("sol", h.profile(), "the cool-off expired on the injected clock — sol is spawnable again");
+        assertEquals(1, adapter.spawnCount("sol"));
+    }
+
+    /**
+     * A credential that is BOTH exhaustion-quarantined (CB-578 stage B) and cooling off (fleetd
+     * #201 Unit 5) at once — the two checks are independent and can both be true for one profile.
+     * Exhaustion quarantine must win: the refusal names quarantine, never cooling off, because
+     * {@code enforceNotQuarantined} is checked (and throws) before {@code enforceNotCoolingOff} ever
+     * runs.
+     */
+    @Test
+    void quarantineWinsOverCoolingOffWhenBothAreActiveOnTheSameCredential() {
+        FakeHerdr herdr = new FakeHerdr();
+        Map<String, FleetConfig.Profile> profiles = ordered(
+                "sol", stubWorker("sol", "shared-openai"),
+                "terra", stubWorker("terra", "shared-openai"));
+        StubLauncher adapter = new StubLauncher("claude", herdr, profiles, "sol", Set.of());
+        BackendQuarantine quarantine = new BackendQuarantine(() -> 0L, TimeUnit.MINUTES.toNanos(30));
+        quarantine.quarantine("shared-openai");
+        BackendOutagePolicy outagePolicy = new BackendOutagePolicy(() -> 0L);
+        outagePolicy.record("shared-openai", "term-1", "API Error: 500");
+        outagePolicy.record("shared-openai", "term-2", "API Error: 500");
+        CompositePeerLauncher composite = new CompositePeerLauncher(List.of(adapter), "sol", profiles,
+                PlacementPolicies.fixed(), _ -> 0, null, quarantine, outagePolicy);
+
+        PlacementException e = assertThrows(PlacementException.class,
+                () -> composite.spawn(new SpawnRequest("sol", null, null)));
+        assertTrue(e.getMessage().contains("quarantined"),
+                "exhaustion quarantine takes priority in the message: " + e.getMessage());
+        assertFalse(e.getMessage().contains("cooling off"),
+                "cooling off is never mentioned when quarantine already refused the spawn: " + e.getMessage());
+    }
+
+    @Test
+    void aFleetWithNoBackendErrorsEverRecordedBehavesExactlyAsBeforeCoolOffExisted() {
+        // A never-record()-called BackendOutagePolicy is naturally inert — the 2/5/6-arg
+        // constructors used throughout this file all wire this in implicitly (NO_OUTAGE_POLICY).
+        // This test pins that an explicit fresh instance also never refuses a spawn, for any profile.
+        FakeHerdr herdr = new FakeHerdr();
+        PeerLauncher composite = composite(herdr);
+
+        assertDoesNotThrow(() -> composite.spawn(new SpawnRequest("claude", null, null)));
+        assertDoesNotThrow(() -> composite.spawn(new SpawnRequest("gemini", null, null)));
+    }
 }

fleetd/src/test/java/dev/ltms/fleet/member/MemberCredentialPolicyViewTest.java

@@ -0,0 +1,99 @@
+package dev.ltms.fleet.member;
+
+import dev.ltms.fleet.config.FleetConfig;
+import org.junit.jupiter.api.Test;
+
+import java.util.List;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertFalse;
+import static org.junit.jupiter.api.Assertions.assertTrue;
+
+/**
+ * fleetd #111 (CB-608): {@link MemberCredentialPolicyView} is the one place that turns a {@code
+ * memberCredentials:} policy into names-and-counts, so the startup log line and {@code
+ * GET /member-credentials} cannot drift apart. These tests pin: the counts always match the
+ * policy that produced them, an absent/empty policy is represented honestly (never as "nothing
+ * blocked"), and the view carries names only — no value ever flows through it, because it is
+ * built only from {@link FleetConfig.MemberCredentials}, which itself never holds a value.
+ */
+class MemberCredentialPolicyViewTest {
+
+    @Test
+    void nullPolicyIsAbsentNotClean() {
+        MemberCredentialPolicyView view = MemberCredentialPolicyView.of(null);
+
+        assertFalse(view.present(), "a null policy must be reported as absent");
+        assertEquals(0, view.knownCount());
+        assertEquals(0, view.allowedCount());
+        assertEquals(0, view.blockedCount());
+        assertTrue(view.known().isEmpty());
+        assertTrue(view.allowed().isEmpty());
+        assertTrue(view.blocked().isEmpty());
+    }
+
+    @Test
+    void emptyKnownListIsAbsentEvenWithAPolicyModeSet() {
+        // A memberCredentials: block can be present in YAML with policy: set but known: empty —
+        // that must still read as "no policy configured", the same as a fully absent block,
+        // because zero known names means the daemon blocks nothing either way.
+        FleetConfig.MemberCredentials creds =
+                new FleetConfig.MemberCredentials("deny-by-default", List.of(), List.of());
+
+        MemberCredentialPolicyView view = MemberCredentialPolicyView.of(creds);
+
+        assertFalse(view.present());
+        assertEquals(0, view.knownCount());
+    }
+
+    @Test
+    void countsMatchARealPolicyExactly() {
+        FleetConfig.MemberCredentials creds = new FleetConfig.MemberCredentials(
+                "deny-by-default",
+                List.of("AI_GATEWAY_TOKEN"),
+                List.of("AI_GATEWAY_TOKEN", "GITEA_ACCESS_TOKEN", "WORKER_GITEA_TOKEN"));
+
+        MemberCredentialPolicyView view = MemberCredentialPolicyView.of(creds);
+
+        assertTrue(view.present());
+        assertEquals("deny-by-default", view.policy());
+        assertEquals(List.of("AI_GATEWAY_TOKEN", "GITEA_ACCESS_TOKEN", "WORKER_GITEA_TOKEN"), view.known());
+        assertEquals(List.of("AI_GATEWAY_TOKEN"), view.allowed());
+        assertEquals(3, view.knownCount());
+        assertEquals(1, view.allowedCount());
+        // known minus allowed — the two names actually shadowed on a spawn.
+        assertEquals(2, view.blockedCount());
+        assertTrue(view.blocked().containsAll(List.of("GITEA_ACCESS_TOKEN", "WORKER_GITEA_TOKEN")));
+    }
+
+    @Test
+    void presentPolicyThatBlocksNothingIsStillDistinctFromAbsent() {
+        // known == allow => blockedCount is 0, exactly like an absent policy's blockedCount — the
+        // two must still be told apart by `present`, or a reader cannot tell "policy configured,
+        // nothing currently blocked" from "no policy at all".
+        FleetConfig.MemberCredentials creds = new FleetConfig.MemberCredentials(
+                "deny-by-default", List.of("X"), List.of("X"));
+
+        MemberCredentialPolicyView view = MemberCredentialPolicyView.of(creds);
+
+        assertTrue(view.present());
+        assertEquals(1, view.knownCount());
+        assertEquals(0, view.blockedCount());
+        assertFalse(MemberCredentialPolicyView.absent().present());
+    }
+
+    @Test
+    void namesPassThroughUnchangedNeverAValue() {
+        // The view is built only from FleetConfig.MemberCredentials, which itself carries names,
+        // never values (see its javadoc) — so there is no code path here that could substitute a
+        // secret's value for its name. This pins the identity: what goes into `known`/`allow` is
+        // exactly what comes out, character for character.
+        List<String> known = List.of("SOME_TOKEN_NAME", "ANOTHER_NAME");
+        FleetConfig.MemberCredentials creds =
+                new FleetConfig.MemberCredentials("deny-by-default", List.of(), known);
+
+        MemberCredentialPolicyView view = MemberCredentialPolicyView.of(creds);
+
+        assertEquals(known, view.known());
+    }
+}

fleetd/src/test/java/dev/ltms/fleet/member/OpenCodeLauncherTest.java

@@ -323,11 +323,39 @@ class OpenCodeLauncherTest {
 
     // --- CB-547: resume + post-hoc session discovery --------------------------------------------
 
+    /**
+     * Give {@code dir} the exact signature {@link HerdrPeerLauncher#isProvisionedWorktree} checks
+     * for: a {@code .git} REGULAR FILE, never a directory. Content is never parsed by that gate, so
+     * any {@code gitdir:} pointer is fine. Mirrors {@code ClaudeCodeLauncherTest}'s helper of the
+     * same shape (fleetd #249).
+     */
+    private static void markAsProvisionedWorktree(Path dir) throws IOException {
+        Files.writeString(dir.resolve(".git"), "gitdir: /tmp/not-a-real-gitdir");
+    }
+
+    /**
+     * A fresh subdirectory of {@code configRoot}, marked as a provisioned worktree (fleetd #249),
+     * for tests that predate this gate and stood in a bare {@code "/work/dir"} string as their
+     * member's cwd — a directory that never existed on disk and, post-#249, would never pass
+     * {@link HerdrPeerLauncher#isProvisionedWorktree} either. Those tests are about the model
+     * mismatch / late-resolve machinery (fleetd #175/#234/#209), not about the worktree gate
+     * itself, so they need a cwd the gate accepts without changing what each test demonstrates.
+     */
+    private static String provisionedWorkDir(Path configRoot) throws IOException {
+        Path dir = Files.createDirectories(configRoot.resolve("work-dir"));
+        markAsProvisionedWorktree(dir);
+        return dir.toString();
+    }
+
     @Test
-    void aResumeSpawnPassesTheSessionIdAsDashS(@TempDir Path root) {
+    void aResumeSpawnIntoAProvisionedWorktreePassesTheSessionIdAsDashS(@TempDir Path root,
+                                                                        @TempDir Path worktree)
+            throws Exception {
+        markAsProvisionedWorktree(worktree);
         FakeHerdr herdr = new FakeHerdr();
         service(herdr, root, opencodeCfg("google/gemini-2.5-pro", null, null))
-                .spawn(new SpawnRequest(null, null, null, null, "ses_41b79fc90ffeI9E8uZv6VprUn2"));
+                .spawn(new SpawnRequest(null, worktree.toString(), null, null,
+                        "ses_41b79fc90ffeI9E8uZv6VprUn2"));
 
         List<String> args = startArgs(herdr);
         int s = args.indexOf("-s");
@@ -336,6 +364,27 @@ class OpenCodeLauncherTest {
                 "the resume target id follows -s");
     }
 
+    /**
+     * fleetd #249 acceptance criterion 3: without a fleetd-provisioned worktree, the member's cwd
+     * is shared with other sessions, so fleetd can never reliably confirm (now or later via {@link
+     * OpenCodeSessionDiscovery}) which conversation it is actually running. Refuse the spawn itself
+     * rather than silently launching opencode's {@code -s <id>} into unverifiable territory.
+     */
+    @Test
+    void aResumeSpawnWithoutAProvisionedWorktreeIsRefused(@TempDir Path root) {
+        FakeHerdr herdr = new FakeHerdr();
+        OpenCodeLauncher launcher = service(herdr, root,
+                opencodeCfg("google/gemini-2.5-pro", null, null));
+
+        IllegalArgumentException e = assertThrows(IllegalArgumentException.class, () ->
+                launcher.spawn(new SpawnRequest(null, null, null, null,
+                        "ses_41b79fc90ffeI9E8uZv6VprUn2")));
+
+        assertTrue(e.getMessage().contains("worktree"), e.getMessage());
+        assertFalse(herdr.called("agent.start"),
+                "the refusal must happen before anything spawns — no pane, no process");
+    }
+
     @Test
     void aFreshSpawnCarriesNoSessionFlag(@TempDir Path root) {
         FakeHerdr herdr = new FakeHerdr();
@@ -348,25 +397,59 @@ class OpenCodeLauncherTest {
 
     @Test
     void theHandleDiscoversTheSessionIdForTheWorkersCwdOnlyAfterItAppears(@TempDir Path root,
-                                                                          @TempDir Path discRoot)
+                                                                          @TempDir Path discRoot,
+                                                                          @TempDir Path worktree)
             throws Exception {
+        markAsProvisionedWorktree(worktree);
         FakeHerdr herdr = new FakeHerdr();
         OpenCodeLauncher launcher = new OpenCodeLauncher(new AgentControl(herdr),
                 new WorkspaceControl(herdr), Map.of("gemini", opencodeCfg(null, null, null)),
                 "gemini", _ -> null, 0, System::currentTimeMillis, () -> { }, root, discRoot);
 
-        PeerHandle handle = launcher.spawn(new SpawnRequest(null, "/work/dir", null));
+        PeerHandle handle = launcher.spawn(new SpawnRequest(null, worktree.toString(), null));
 
         // opencode writes the record only when the session is first persisted — the instant the
         // pane is ready it does not exist, so agentSessionId() is null (never a spawn failure).
         assertNull(handle.agentSessionId(), "no record yet → null, not a spawn-time block");
         // Once the record appears (here: same cwd), lazy discovery resolves it — the handle's
         // session id matches its own worktree, not another's.
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_resolved", "/work/dir", 1000L);
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_resolved", worktree.toString(), 1000L);
         assertEquals("ses_resolved", handle.agentSessionId(),
                 "agentSessionId() re-scans and picks up a record that has since been written");
     }
 
+    /**
+     * fleetd #249 acceptance criterion 1, exercised through the real caller path (the handle
+     * {@code fleet_list} actually reads), not {@link OpenCodeSessionDiscovery} directly. Without a
+     * fleetd-provisioned worktree the member's cwd is shared — the default no-worktree spawn
+     * inherits the lead's own long-lived cwd — so even once a matching row appears (here:
+     * simulating another profile's session that happens to share the directory) the handle must
+     * report absence rather than guess. Measured real-world case (2026-09-03): the row it would
+     * otherwise pick was three days old and belonged to a different profile.
+     */
+    @Test
+    void theHandleNeverReportsAnIdForANonProvisionedCwdEvenAfterARowAppears(@TempDir Path root,
+                                                                             @TempDir Path discRoot)
+            throws Exception {
+        FakeHerdr herdr = new FakeHerdr();
+        OpenCodeLauncher launcher = new OpenCodeLauncher(new AgentControl(herdr),
+                new WorkspaceControl(herdr), Map.of("gemini", opencodeCfg(null, null, null)),
+                "gemini", _ -> null, 0, System::currentTimeMillis, () -> { }, root, discRoot);
+        // No markAsProvisionedWorktree — this cwd has no .git file, the shared-cwd shape a
+        // no-worktree spawn (or a real checkout) actually has.
+        String sharedCwd = root.resolve("shared-cwd").toString();
+
+        PeerHandle handle = launcher.spawn(new SpawnRequest(null, sharedCwd, null));
+
+        assertNull(handle.agentSessionId(), "no record yet → null, same as the provisioned case");
+        // A row for this exact directory now appears — e.g. a sibling member, or a stale session
+        // from days earlier, sharing the same unprovisioned cwd.
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_someone_elses", sharedCwd, 1000L);
+        assertNull(handle.agentSessionId(),
+                "a non-provisioned cwd must NEVER report an id, even once a row for it exists — "
+                        + "the row could belong to any other session sharing this directory");
+    }
+
     @Test
     void foreignWorkerMatchesOpencodePrefixButNotClaude() {
         String nonce = "abc123";
@@ -920,6 +1003,7 @@ class OpenCodeLauncherTest {
     @Test
     void theRealSessionManagerLateResolvePathCatchesAModelMismatch(@TempDir Path configRoot,
                                                                      @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         FakeHerdr herdr = new FakeHerdr();
         // xf's real shape (fleetd #175): weight:80, model "opencode/nemotron-3-ultra-free", no
         // credentialId — the profile that actually escaped the fleet's accounting.
@@ -929,7 +1013,7 @@ class OpenCodeLauncherTest {
         OpenCodeLauncher launcher = serviceWithSink(herdr, configRoot, discRoot, cfg, sink);
 
         SessionManager sessions = new SessionManager(launcher);
-        MemberSession acquired = sessions.acquire(cfg.profile(), "/work/dir", null, null);
+        MemberSession acquired = sessions.acquire(cfg.profile(), workDir, null, null);
 
         // Real late-resolve path, driven BEFORE opencode has written its session row — same shape
         // as production the instant a pane goes ready.
@@ -940,7 +1024,7 @@ class OpenCodeLauncherTest {
 
         // opencode writes its row late, running gpt-5.6-sol (a PAID credential) instead of the
         // withdrawn free model the profile actually asked for — the exact fleetd #175 scenario.
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"gpt-5.6-sol\",\"providerID\":\"openai\"}");
 
         // Drive the SAME real late-resolve path again: sessions.get() -> resolveAgentSessionId ->
@@ -959,12 +1043,13 @@ class OpenCodeLauncherTest {
     @Test
     void aProviderPrefixedModelMatchingBothIdAndProviderIsNotAMismatch(@TempDir Path configRoot,
                                                                         @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         List<String> exhausted = new ArrayList<>();
         ExhaustionSink sink = (target, reason, profile) -> exhausted.add(reason);
         FleetConfig.Profile cfg = opencodeCfg("openai/gpt-5.6-terra", null, null);
         PeerHandle handle = serviceWithSink(new FakeHerdr(), configRoot, discRoot, cfg, sink)
-                .spawn(new SpawnRequest(null, "/work/dir", null));
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+                .spawn(new SpawnRequest(null, workDir, null));
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"gpt-5.6-terra\",\"providerID\":\"openai\"}");
 
         assertEquals("ses_x", handle.agentSessionId());
@@ -975,12 +1060,13 @@ class OpenCodeLauncherTest {
     @Test
     void aGxProviderPrefixedModelMatchingBothIdAndProviderIsNotAMismatch(@TempDir Path configRoot,
                                                                           @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         List<String> exhausted = new ArrayList<>();
         ExhaustionSink sink = (target, reason, profile) -> exhausted.add(reason);
         FleetConfig.Profile cfg = opencodeCfg("gx/deepseek-v4-flash", null, null);
         PeerHandle handle = serviceWithSink(new FakeHerdr(), configRoot, discRoot, cfg, sink)
-                .spawn(new SpawnRequest(null, "/work/dir", null));
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+                .spawn(new SpawnRequest(null, workDir, null));
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"deepseek-v4-flash\",\"providerID\":\"gx\"}");
 
         assertEquals("ses_x", handle.agentSessionId());
@@ -997,12 +1083,13 @@ class OpenCodeLauncherTest {
     @Test
     void aMissingProviderIdInTheEvidenceIsUnknownNotAMismatchWhenTheIdMatches(
             @TempDir Path configRoot, @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         List<String> exhausted = new ArrayList<>();
         ExhaustionSink sink = (target, reason, profile) -> exhausted.add(reason);
         FleetConfig.Profile cfg = opencodeCfg("openai/gpt-5.6-terra", null, null);
         PeerHandle handle = serviceWithSink(new FakeHerdr(), configRoot, discRoot, cfg, sink)
-                .spawn(new SpawnRequest(null, "/work/dir", null));
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+                .spawn(new SpawnRequest(null, workDir, null));
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"gpt-5.6-terra\"}");
 
         assertEquals("ses_x", handle.agentSessionId());
@@ -1018,12 +1105,13 @@ class OpenCodeLauncherTest {
     @Test
     void aMissingProviderIdInTheEvidenceStillCatchesARealIdMismatch(
             @TempDir Path configRoot, @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         List<String> exhausted = new ArrayList<>();
         ExhaustionSink sink = (target, reason, profile) -> exhausted.add(reason);
         FleetConfig.Profile cfg = opencodeCfg("openai/gpt-5.6-terra", null, null);
         PeerHandle handle = serviceWithSink(new FakeHerdr(), configRoot, discRoot, cfg, sink)
-                .spawn(new SpawnRequest(null, "/work/dir", null));
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+                .spawn(new SpawnRequest(null, workDir, null));
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"gpt-5.6-sol\"}");
 
         assertEquals("ses_x", handle.agentSessionId());
@@ -1042,12 +1130,13 @@ class OpenCodeLauncherTest {
     @Test
     void aBareModelWithNoProviderPrefixMatchesOnIdAloneAndIsNotAMismatch(@TempDir Path configRoot,
                                                                           @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         List<String> exhausted = new ArrayList<>();
         ExhaustionSink sink = (target, reason, profile) -> exhausted.add(reason);
         FleetConfig.Profile cfg = opencodeCfg("deepseek-v4-flash", null, null);
         PeerHandle handle = serviceWithSink(new FakeHerdr(), configRoot, discRoot, cfg, sink)
-                .spawn(new SpawnRequest(null, "/work/dir", null));
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+                .spawn(new SpawnRequest(null, workDir, null));
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"deepseek-v4-flash\",\"providerID\":\"gx\"}");
 
         assertEquals("ses_x", handle.agentSessionId());
@@ -1064,6 +1153,7 @@ class OpenCodeLauncherTest {
     @Test
     void aRealIdMismatchLogsAnErrorNamingBothModelsAndQuarantinesThroughTheSink(
             @TempDir Path configRoot, @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         List<String> exhausted = new ArrayList<>();
         ExhaustionSink sink = (target, reason, profile) -> exhausted.add(target + "|" + reason);
         FleetConfig.Profile cfg = opencodeCfg("opencode/nemotron-3-ultra-free", null, null);
@@ -1075,8 +1165,8 @@ class OpenCodeLauncherTest {
         PeerHandle handle;
         try {
             handle = serviceWithSink(new FakeHerdr(), configRoot, discRoot, cfg, sink)
-                    .spawn(new SpawnRequest(null, "/work/dir", null));
-            OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+                    .spawn(new SpawnRequest(null, workDir, null));
+            OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                     "{\"id\":\"gpt-5.6-sol\",\"providerID\":\"openai\"}");
             assertEquals("ses_x", handle.agentSessionId());
         } finally {
@@ -1111,11 +1201,12 @@ class OpenCodeLauncherTest {
     @Test
     void unknownOrUnparseableModelEvidenceNeverQuarantines(@TempDir Path configRoot,
                                                              @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         List<String> exhausted = new ArrayList<>();
         ExhaustionSink sink = (target, reason, profile) -> exhausted.add(reason);
         FleetConfig.Profile cfg = opencodeCfg("openai/gpt-5.6-terra", null, null);
         PeerHandle handle = serviceWithSink(new FakeHerdr(), configRoot, discRoot, cfg, sink)
-                .spawn(new SpawnRequest(null, "/work/dir", null));
+                .spawn(new SpawnRequest(null, workDir, null));
 
         // No row yet at all.
         assertNull(handle.agentSessionId());
@@ -1137,12 +1228,13 @@ class OpenCodeLauncherTest {
     @Test
     void aProfileWithNoConfiguredModelIsNeverCheckedForAMismatch(@TempDir Path configRoot,
                                                                   @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         List<String> exhausted = new ArrayList<>();
         ExhaustionSink sink = (target, reason, profile) -> exhausted.add(reason);
         FleetConfig.Profile cfg = opencodeCfg(null, null, null);
         PeerHandle handle = serviceWithSink(new FakeHerdr(), configRoot, discRoot, cfg, sink)
-                .spawn(new SpawnRequest(null, "/work/dir", null));
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+                .spawn(new SpawnRequest(null, workDir, null));
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"anything-at-all\",\"providerID\":\"anyone\"}");
 
         assertEquals("ses_x", handle.agentSessionId());
@@ -1166,21 +1258,22 @@ class OpenCodeLauncherTest {
     @Test
     void modelCheckReadsTheResolvedSessionsOwnRowNotWhateverIsNewestInTheSharedDirectory(
             @TempDir Path configRoot, @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         List<String> exhausted = new ArrayList<>();
         ExhaustionSink sink = (target, reason, profile) -> exhausted.add(reason);
         FleetConfig.Profile cfg = opencodeCfg("openai/gpt-5.6-terra", null, null);
         PeerHandle handle = serviceWithSink(new FakeHerdr(), configRoot, discRoot, cfg, sink)
-                .spawn(new SpawnRequest(null, "/work/dir", null));
+                .spawn(new SpawnRequest(null, workDir, null));
 
         // Our own session's row, correctly matching the profile's requested model.
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_ours", "/work/dir", 1000L,
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_ours", workDir, 1000L,
                 "{\"id\":\"gpt-5.6-terra\",\"providerID\":\"openai\"}");
         assertEquals("ses_ours", handle.agentSessionId(), "resolves to our own session");
         assertTrue(exhausted.isEmpty(), "matching model → no mismatch on first resolve: " + exhausted);
 
         // A sibling member, spawned later into the SAME shared directory (no worktree, fleetd
         // #234's default), writes a newer row running a totally different model.
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_sibling", "/work/dir", 9000L,
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_sibling", workDir, 9000L,
                 "{\"id\":\"deepseek-v4-flash\",\"providerID\":\"gx\"}");
 
         assertEquals("ses_ours", handle.agentSessionId(),
@@ -1212,6 +1305,7 @@ class OpenCodeLauncherTest {
     @Test
     void aSpawnTimeModelMismatchActuallyQuarantinesTheCredentialThroughTheRealAcquirePath(
             @TempDir Path configRoot, @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         FleetConfig.Profile cfg = opencodeCfgWithCredential(
                 "terra", "opencode/nemotron-3-ultra-free", "openai-shared");
         Map<String, FleetConfig.Profile> profiles = Map.of(cfg.profile(), cfg);
@@ -1232,14 +1326,14 @@ class OpenCodeLauncherTest {
 
         // The mismatching row exists BEFORE the spawn — reproducing fleetd #234's exact timing:
         // opencode's session table already carries evidence by the moment acquire() first asks.
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"gpt-5.6-sol\",\"providerID\":\"openai\"}");
 
         assertFalse(quarantine.isQuarantined("openai-shared"), "nothing quarantined before the spawn");
 
         // The real production entrypoint: acquire() builds the MemberSession by calling
         // handle.agentSessionId() BEFORE registry.put() runs.
-        MemberSession acquired = sessions.acquire(cfg.profile(), "/work/dir", null, null);
+        MemberSession acquired = sessions.acquire(cfg.profile(), workDir, null, null);
 
         assertEquals("ses_x", acquired.agentSessionId(), "the id itself still resolves correctly");
         assertTrue(quarantine.isQuarantined("openai-shared"),
@@ -1258,6 +1352,7 @@ class OpenCodeLauncherTest {
     @Test
     void aRosterOnlySinkSilentlyDropsTheSpawnTimeQuarantine(@TempDir Path configRoot,
                                                               @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         FleetConfig.Profile cfg = opencodeCfgWithCredential(
                 "terra", "opencode/nemotron-3-ultra-free", "openai-shared");
         BackendQuarantine quarantine = new BackendQuarantine(() -> 0L, TimeUnit.SECONDS.toNanos(1800));
@@ -1270,10 +1365,10 @@ class OpenCodeLauncherTest {
         OpenCodeLauncher launcher = serviceWithSink(herdr, configRoot, discRoot, cfg, rosterOnlySink);
         SessionManager sessions = new SessionManager(launcher);
 
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"gpt-5.6-sol\",\"providerID\":\"openai\"}");
 
-        MemberSession acquired = sessions.acquire(cfg.profile(), "/work/dir", null, null);
+        MemberSession acquired = sessions.acquire(cfg.profile(), workDir, null, null);
 
         assertEquals("ses_x", acquired.agentSessionId(), "the id itself still resolves correctly");
         assertFalse(quarantine.isQuarantined("openai-shared"),
@@ -1301,6 +1396,7 @@ class OpenCodeLauncherTest {
     @Test
     void theSpawnTimeQuarantineSurvivesTheFleetdStyleForwardingHop(@TempDir Path configRoot,
                                                                      @TempDir Path discRoot) throws Exception {
+        String workDir = provisionedWorkDir(configRoot);
         FleetConfig.Profile cfg = opencodeCfgWithCredential(
                 "terra", "opencode/nemotron-3-ultra-free", "openai-shared");
         Map<String, FleetConfig.Profile> profiles = Map.of(cfg.profile(), cfg);
@@ -1332,12 +1428,12 @@ class OpenCodeLauncherTest {
         };
         exhaustionSinkRef.set(realSink);
 
-        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", "/work/dir", 1000L,
+        OpenCodeSessionDiscoveryTest.writeRecord(discRoot, "ses_x", workDir, 1000L,
                 "{\"id\":\"gpt-5.6-sol\",\"providerID\":\"openai\"}");
 
         assertFalse(quarantine.isQuarantined("openai-shared"), "nothing quarantined before the spawn");
 
-        MemberSession acquired = sessions.acquire(cfg.profile(), "/work/dir", null, null);
+        MemberSession acquired = sessions.acquire(cfg.profile(), workDir, null, null);
 
         assertEquals("ses_x", acquired.agentSessionId(), "the id itself still resolves correctly");
         assertTrue(quarantine.isQuarantined("openai-shared"),

fleetd/src/test/java/dev/ltms/fleet/placement/PlacementPolicyTest.java

@@ -30,7 +30,14 @@ class PlacementPolicyTest {
     private static PlacementContext ctx(List<PlacementCandidate> candidates,
                                         Function<String, Integer> liveCount,
                                         Set<String> unreachable, Set<String> quarantined) {
-        return new PlacementContext("b", candidates, liveCount, unreachable, quarantined);
+        return ctx(candidates, liveCount, unreachable, quarantined, Set.of());
+    }
+
+    private static PlacementContext ctx(List<PlacementCandidate> candidates,
+                                        Function<String, Integer> liveCount,
+                                        Set<String> unreachable, Set<String> quarantined,
+                                        Set<String> coolingOff) {
+        return new PlacementContext("b", candidates, liveCount, unreachable, quarantined, coolingOff);
     }
 
     private static PlacementContext ctx(List<PlacementCandidate> candidates,
@@ -52,14 +59,14 @@ class PlacementPolicyTest {
         PlacementPolicy policy = PlacementPolicies.fixed();
         PlacementContext ctx = new PlacementContext(null,
                 List.of(PlacementCandidate.profile("a"), PlacementCandidate.profile("b")),
-                noSessions(), Set.of(), Set.of());
+                noSessions(), Set.of(), Set.of(), Set.of());
         assertEquals("a", policy.select(ctx).profile());
     }
 
     @Test
     void fixedThrowsWhenNoProfilesAndNoDefault() {
         PlacementPolicy policy = PlacementPolicies.fixed();
-        PlacementContext ctx = new PlacementContext(null, List.of(), noSessions(), Set.of(), Set.of());
+        PlacementContext ctx = new PlacementContext(null, List.of(), noSessions(), Set.of(), Set.of(), Set.of());
         assertThrows(PlacementException.class, () -> policy.select(ctx));
     }
 
@@ -68,7 +75,7 @@ class PlacementPolicyTest {
         PlacementPolicy policy = PlacementPolicies.fixed();
         PlacementContext ctx = new PlacementContext("b",
                 List.of(PlacementCandidate.profile("a"), PlacementCandidate.profile("b")),
-                noSessions(), Set.of(), Set.of("b"));
+                noSessions(), Set.of(), Set.of("b"), Set.of());
         assertEquals("a", policy.select(ctx).profile(),
                 "the default 'b' is quarantined, so fixed falls through to the first un-quarantined candidate");
     }
@@ -78,7 +85,7 @@ class PlacementPolicyTest {
         PlacementPolicy policy = PlacementPolicies.fixed();
         PlacementContext ctx = new PlacementContext("b",
                 List.of(PlacementCandidate.profile("a"), PlacementCandidate.profile("b")),
-                noSessions(), Set.of(), Set.of("a", "b"));
+                noSessions(), Set.of(), Set.of("a", "b"), Set.of());
         PlacementException e = assertThrows(PlacementException.class, () -> policy.select(ctx));
         assertTrue(e.getMessage().contains("quarantined"), e.getMessage());
     }
@@ -211,6 +218,67 @@ class PlacementPolicyTest {
         assertTrue(e.getMessage().contains("quarantined"), e.getMessage());
     }
 
+    // --- fleetd #201 Unit 5: coolingOff excludes a candidate, SEPARATE from quarantined --------
+
+    @Test
+    void roundRobinSkipsCoolingOffProfiles() {
+        PlacementPolicy policy = PlacementPolicies.roundRobin();
+        List<PlacementCandidate> candidates = List.of(
+                PlacementCandidate.profile("a"),
+                PlacementCandidate.profile("b"));
+        PlacementContext ctx = ctx(candidates, noSessions(), Set.of(), Set.of(), Set.of("a"));
+        for (int i = 0; i < 5; i++) {
+            assertEquals("b", policy.select(ctx).profile(), "a is cooling off, so every pick lands on b");
+        }
+    }
+
+    @Test
+    void weightedSkipsCoolingOffProfile() {
+        PlacementPolicy policy = PlacementPolicies.weighted();
+        List<PlacementCandidate> candidates = List.of(
+                PlacementCandidate.profile("a", 1.0f, null),
+                PlacementCandidate.profile("b", 1.0f, null));
+        PlacementContext ctx = ctx(candidates, noSessions(), Set.of(), Set.of(), Set.of("a"));
+        for (int i = 0; i < 5; i++) {
+            assertEquals("b", policy.select(ctx).profile(), "a is cooling off, so every pick lands on b");
+        }
+    }
+
+    @Test
+    void weightedThrowsWhenAllCoolingOff() {
+        PlacementPolicy policy = PlacementPolicies.weighted();
+        List<PlacementCandidate> candidates = List.of(
+                PlacementCandidate.profile("a"),
+                PlacementCandidate.profile("b"));
+        PlacementException e = assertThrows(PlacementException.class,
+                () -> policy.select(ctx(candidates, noSessions(), Set.of(), Set.of(), Set.of("a", "b"))));
+        assertTrue(e.getMessage().contains("cooling off"), e.getMessage());
+        assertFalse(e.getMessage().contains("quarantined"),
+                "nothing here is quarantined — the message must say cooling off, not exhausted: " + e.getMessage());
+    }
+
+    /**
+     * A candidate that is BOTH quarantined (CB-578 stage B) and cooling off (fleetd #201 Unit 5)
+     * counts only as quarantined — exhaustion takes priority, matching
+     * {@code CompositePeerLauncher}'s explicit-spawn check order.
+     */
+    @Test
+    void quarantinedAndCoolingOffIsNotDoubleCounted() {
+        PlacementPolicy policy = PlacementPolicies.weighted();
+        List<PlacementCandidate> candidates = List.of(
+                PlacementCandidate.profile("a"),
+                PlacementCandidate.profile("b"));
+        // a is BOTH quarantined and cooling off; b is quarantined only. If a were double-bucketed
+        // as cooling-off instead of quarantined, quarantined would undercount to 1 of 2 candidates
+        // and this would fall through to the generic "no worker profile available: N quarantined, M
+        // cooling off, ..." message instead — which also happens to contain the substring
+        // "quarantined", so a loose contains() check here would pass either way. Pinning the exact
+        // "all ... quarantined" message is what actually proves the two are not double-counted.
+        PlacementContext ctx = ctx(candidates, noSessions(), Set.of(), Set.of("a", "b"), Set.of("a"));
+        PlacementException e = assertThrows(PlacementException.class, () -> policy.select(ctx));
+        assertEquals("all worker profiles are quarantined (backend exhausted)", e.getMessage());
+    }
+
     @Test
     void weightedThrowsWhenAllUnreachable() {
         PlacementPolicy policy = PlacementPolicies.weighted();
@@ -320,7 +388,7 @@ class PlacementPolicyTest {
         PlacementContext ctx = new PlacementContext("b",
                 List.of(PlacementCandidate.profile("a", 1.0f, null),
                         PlacementCandidate.profile("b", 0.0f, null)),
-                noSessions(), Set.of(), Set.of());
+                noSessions(), Set.of(), Set.of(), Set.of());
         assertEquals("a", policy.select(ctx).profile(),
                 "the default 'b' has weight 0, so fixed falls through to the first available candidate");
     }
@@ -331,7 +399,7 @@ class PlacementPolicyTest {
         PlacementContext ctx = new PlacementContext("b",
                 List.of(PlacementCandidate.profile("a", 0.0f, null),
                         PlacementCandidate.profile("b", 0.0f, null)),
-                noSessions(), Set.of(), Set.of());
+                noSessions(), Set.of(), Set.of(), Set.of());
         PlacementException e = assertThrows(PlacementException.class, () -> policy.select(ctx));
         assertTrue(e.getMessage().contains("weight 0"), e.getMessage());
     }

fleetd/src/test/java/dev/ltms/fleet/session/GitWorktreesTest.java

@@ -666,6 +666,103 @@ class GitWorktreesTest {
         assertEquals("", status(Path.of(wt), ".autoenv"), ".autoenv still shows as modified");
     }
 
+    // ---- fleetd #134: isolateToolSurface must report what it neutralized (to the daemon operator's
+    // log) and record it where the worker itself can read it (worktree-scoped git config), without
+    // ever showing up in the worker's own `git status`. All drive the real provisioning path,
+    // GitWorktrees#add, per criterion 5 — the whole provisioned worktree is what's under test here. ----
+
+    private static Path initRepoWithAllThreeConfigs(Path dir) throws Exception {
+        Files.createDirectories(dir);
+        git(dir, "init", "-q", "-b", "main");
+        git(dir, "config", "user.email", "test@example.invalid");
+        git(dir, "config", "user.name", "Test");
+        Files.writeString(dir.resolve(".mcp.json"), WITH_SERVERS);
+        Files.writeString(dir.resolve("opencode.json"), OPENCODE_WITH_FILE_REF);
+        Files.writeString(dir.resolve(".autoenv"), AUTOENV_WITH_DIRECTIVE);
+        Files.writeString(dir.resolve("README.md"), "seed\n");
+        git(dir, "add", ".mcp.json", "opencode.json", ".autoenv", "README.md");
+        git(dir, "commit", "-q", "-m", "seed");
+        return dir;
+    }
+
+    /** Criterion 1, all three present: the summary names the denominator and every neutralized file. */
+    @Test
+    void isolateToolSurfaceLogsAllThreeConfigsNeutralized(@TempDir Path tmp) throws Exception {
+        reportingLogger.setLevel(Level.INFO);
+        Path repo = initRepoWithAllThreeConfigs(tmp.resolve("repo"));
+
+        new GitWorktrees(tmp.resolve("wts").toString()).add(repo.toString(), "cb-134-log-all", "HEAD");
+
+        assertTrue(capturedMessages().contains(
+                        "tool-surface isolation: neutralized 3 of 3 configs: .mcp.json, opencode.json, "
+                                + ".autoenv — the worktree copy is a stub, not the repo's file; edit the "
+                                + "real file in the primary checkout instead"),
+                "expected the all-neutralized summary line, got:\n" + capturedMessages());
+    }
+
+    /** Criterion 1, two absent: the summary must still name the denominator and say why. */
+    @Test
+    void isolateToolSurfaceLogsAbsentConfigsWithReason(@TempDir Path tmp) throws Exception {
+        reportingLogger.setLevel(Level.INFO);
+        Path repo = initRepo(tmp.resolve("repo")); // only .mcp.json + README committed
+
+        new GitWorktrees(tmp.resolve("wts").toString()).add(repo.toString(), "cb-134-log-partial", "HEAD");
+
+        assertTrue(capturedMessages().contains(
+                        "tool-surface isolation: neutralized 1 of 3 configs: .mcp.json (opencode.json "
+                                + "absent, .autoenv absent) — the worktree copy is a stub, not the repo's "
+                                + "file; edit the real file in the primary checkout instead"),
+                "expected the partial summary line, got:\n" + capturedMessages());
+    }
+
+    /**
+     * Criterion 2. The daemon's own log is invisible to the worker process — it never reads fleetd's
+     * stdout. This is the mechanism the worker itself can query, from inside its own worktree, to
+     * learn "this file is neutralized here, the repo's real file differs" instead of trusting what it
+     * just read on disk.
+     */
+    @Test
+    void aWorkerCanDiscoverNeutralizedConfigsFromWorktreeScopedGitConfig(@TempDir Path tmp) throws Exception {
+        Path repo = initRepoWithAllThreeConfigs(tmp.resolve("repo"));
+
+        String wt = new GitWorktrees(tmp.resolve("wts").toString())
+                .add(repo.toString(), "cb-134-discover", "HEAD");
+
+        String recorded = gitOutput(Path.of(wt), "config", "--worktree", "--get-all", "fleet.neutralizedConfig");
+        Set<String> files = new HashSet<>();
+        for (String line : recorded.split("\\R")) {
+            if (!line.isBlank()) {
+                files.add(line.trim());
+            }
+        }
+        assertEquals(Set.of(".mcp.json", "opencode.json", ".autoenv"), files,
+                "the worker-readable record must name every neutralized file: " + recorded);
+
+        String note = gitOutput(Path.of(wt), "config", "--worktree", "--get", "fleet.neutralizedConfigNote").trim();
+        assertTrue(note.contains("stub"), "note must say the worktree copy is a stub: " + note);
+        assertTrue(note.contains("primary checkout"),
+                "note must state the consequence — where to edit the real file instead: " + note);
+    }
+
+    /**
+     * Criterion 3. Whatever fleetd#134's worker-discovery mechanism writes must never appear as
+     * untracked or modified in the worker's own `git status` — a worker that sees a stray file either
+     * commits it by mistake or burns a turn asking about it. This runs the full porcelain status, not
+     * a single-file check, so any leftover file anywhere in the worktree would fail it.
+     */
+    @Test
+    void aProvisionedWorktreeHasCleanGitStatusDespiteNeutralizedConfigRecordkeeping(@TempDir Path tmp)
+            throws Exception {
+        Path repo = initRepoWithAllThreeConfigs(tmp.resolve("repo"));
+
+        String wt = new GitWorktrees(tmp.resolve("wts").toString())
+                .add(repo.toString(), "cb-134-clean-status", "HEAD");
+
+        assertEquals("", fullStatus(Path.of(wt)),
+                "a freshly provisioned worktree must show a clean `git status --porcelain`, including "
+                        + "after the worker-readable neutralized-config record was written");
+    }
+
     /**
      * CB-578 stage C, acceptance criterion 1. A dirty worktree — a tracked edit plus a brand-new
      * untracked file, exactly the shape lost in CB-576 — must land in {@code refs/wip/<branch>}'s
@@ -1177,4 +1274,112 @@ class GitWorktreesTest {
                     + "the refusal must happen before `git worktree add` ever runs");
         }
     }
+
+    // ---- fleetd #134 / #148 point 3: overlayParity must report what it did, and marking a tracked
+    // file --skip-worktree must say the file can no longer be committed from this worktree. Drives
+    // overlayParity directly against a real worktree (git worktree add, no GitWorktrees#add) so these
+    // tests are independent of origin/credential-helper provisioning, which is not under test here. ----
+
+    /** A bare worktree, sibling to {@code repo}, created with plain git — the target overlayParity
+     *  copies into. Deliberately not {@link GitWorktrees#add}: that method does unrelated
+     *  provisioning (origin rewrite, .mcp.json neutralization, credential helper) that would only
+     *  add noise to the log assertions below. */
+    private static Path bareWorktree(Path repo, Path wtDir, String branch) throws Exception {
+        git(repo, "worktree", "add", "-q", wtDir.toString(), "-b", branch, "HEAD");
+        return wtDir;
+    }
+
+    /** Acceptance criterion 1: only the configured candidates land in the worktree — nothing else
+     *  from the source tree leaks in alongside them. */
+    @Test
+    void overlayParityCopiesExactlyTheConfiguredFilesAndNothingElse(@TempDir Path tmp) throws Exception {
+        Path repo = initRepo(tmp.resolve("repo"));
+        Files.writeString(repo.resolve(".env"), "A=1\n");
+        Files.writeString(repo.resolve(".envrc"), "export A=1\n");
+        Files.writeString(repo.resolve("not-overlaid.txt"), "must not be copied\n");
+        Path wt = bareWorktree(repo, tmp.resolve("wt"), "cb134-exact");
+
+        new GitWorktrees(tmp.resolve("wts").toString())
+                .overlayParity(repo.toString(), wt.toString(), List.of(".env", ".envrc"));
+
+        assertEquals("A=1\n", Files.readString(wt.resolve(".env")));
+        assertEquals("export A=1\n", Files.readString(wt.resolve(".envrc")));
+        assertFalse(Files.exists(wt.resolve("not-overlaid.txt")),
+                "overlayParity must copy only the configured candidates, not the whole source tree");
+    }
+
+    /** Criterion 2: both candidates present — the summary line names both and the denominator. */
+    @Test
+    void overlayParityLogsBothCopiedWhenBothCandidatesArePresent(@TempDir Path tmp) throws Exception {
+        reportingLogger.setLevel(Level.INFO);
+        Path repo = initRepo(tmp.resolve("repo"));
+        Files.writeString(repo.resolve(".env"), "A=1\n");
+        Files.writeString(repo.resolve(".envrc"), "export A=1\n");
+        Path wt = bareWorktree(repo, tmp.resolve("wt"), "cb134-both");
+
+        new GitWorktrees(tmp.resolve("wts").toString())
+                .overlayParity(repo.toString(), wt.toString(), List.of(".env", ".envrc"));
+
+        assertTrue(capturedMessages().contains("parity overlay: copied 2 of 2 candidates: .env, .envrc"),
+                "expected the both-copied summary line, got:\n" + capturedMessages());
+    }
+
+    /** Criterion 2: one candidate present, one absent — the summary must name the copied file, the
+     *  denominator, and why the other candidate was not copied. */
+    @Test
+    void overlayParityLogsOneCopiedOneAbsent(@TempDir Path tmp) throws Exception {
+        reportingLogger.setLevel(Level.INFO);
+        Path repo = initRepo(tmp.resolve("repo"));
+        Files.writeString(repo.resolve(".env"), "A=1\n");
+        // .envrc deliberately not created — the absent candidate.
+        Path wt = bareWorktree(repo, tmp.resolve("wt"), "cb134-partial");
+
+        new GitWorktrees(tmp.resolve("wts").toString())
+                .overlayParity(repo.toString(), wt.toString(), List.of(".env", ".envrc"));
+
+        assertTrue(Files.exists(wt.resolve(".env")));
+        assertFalse(Files.exists(wt.resolve(".envrc")));
+        assertTrue(capturedMessages().contains("parity overlay: copied 1 of 2 candidates: .env (.envrc absent)"),
+                "expected the copied/absent summary line, got:\n" + capturedMessages());
+    }
+
+    /** Criterion 3: a tracked candidate is marked --skip-worktree, and that must be named in the log
+     *  with the consequence spelled out — a worker editing it afterward finds git ignoring the
+     *  change, silently, unless this line told it so beforehand. */
+    @Test
+    void overlayParityLogsSkipWorktreeConsequenceForATrackedFile(@TempDir Path tmp) throws Exception {
+        reportingLogger.setLevel(Level.INFO);
+        Path repo = initRepo(tmp.resolve("repo"));
+        Files.writeString(repo.resolve(".env"), "A=1\n");
+        git(repo, "add", ".env");
+        git(repo, "commit", "-q", "-m", "track env");
+        Path wt = bareWorktree(repo, tmp.resolve("wt"), "cb134-tracked");
+        // Change the source after the worktree checkout, so the overlay copy actually overwrites it.
+        Files.writeString(repo.resolve(".env"), "A=2\n");
+
+        new GitWorktrees(tmp.resolve("wts").toString())
+                .overlayParity(repo.toString(), wt.toString(), List.of(".env"));
+
+        assertEquals("A=2\n", Files.readString(wt.resolve(".env")));
+        assertEquals("", status(wt, ".env"),
+                "the skip-worktree'd file must not show as modified even though its content changed");
+        assertTrue(capturedMessages().contains(
+                        "parity overlay marked --skip-worktree (cannot be committed from this worktree): .env"),
+                "expected the skip-worktree consequence line, got:\n" + capturedMessages());
+    }
+
+    /** Criterion 5: null and empty overlay lists return quietly — no exception, no log noise. */
+    @Test
+    void overlayParityWithNoCandidatesLogsNothing(@TempDir Path tmp) throws Exception {
+        reportingLogger.setLevel(Level.INFO);
+        Path repo = initRepo(tmp.resolve("repo"));
+        Path wt = bareWorktree(repo, tmp.resolve("wt"), "cb134-empty");
+        GitWorktrees worktrees = new GitWorktrees(tmp.resolve("wts").toString());
+
+        worktrees.overlayParity(repo.toString(), wt.toString(), null);
+        worktrees.overlayParity(repo.toString(), wt.toString(), List.of());
+
+        assertTrue(reportingAppender.list.isEmpty(),
+                "a null/empty overlay must log nothing, got:\n" + capturedMessages());
+    }
 }

fleetd/src/test/java/dev/ltms/fleet/session/WorktreeSessionManagerTest.java

@@ -140,7 +140,7 @@ class WorktreeSessionManagerTest {
     void worktreeAcquireRunsParityOverlayWithProfileDefaults() {
         FakeHerdr herdr = new FakeHerdr();
         FakeWorktrees worktrees = new FakeWorktrees().withRepoRoot("/repo").withPrefix("/wt")
-                .track(".envrc")
+                .track(".env")
                 .exists(".claude/settings.local.json");
         SessionManager sessions = new SessionManager(workerService(herdr), worktrees);
 
@@ -151,14 +151,15 @@ class WorktreeSessionManagerTest {
         FakeWorktrees.OverlayCall overlay = worktrees.lastOverlay();
         assertNotNull(overlay);
         assertEquals("/repo", overlay.repoRoot());
-        assertEquals(List.of(".env", ".envrc"),
-                overlay.requested(), "default parity overlay is used when unset");
+        assertEquals(List.of(".env"),
+                overlay.requested(), "default parity overlay is used when unset (CB-148: .env only, "
+                        + ".envrc is no longer defaulted because it is executable shell direnv runs on cd)");
         assertFalse(overlay.requested().contains(".mcp.json"),
                 "CB-525: replicating the primary's MCP config gives a worker the primary's IDE "
                         + "servers, which navigate its edits out of its own worktree");
-        assertEquals(List.of(".envrc"), overlay.copied(),
+        assertEquals(List.of(".env"), overlay.copied(),
                 "existing paths are copied; missing paths are skipped");
-        assertEquals(List.of(".envrc"), overlay.skipWorktree(),
+        assertEquals(List.of(".env"), overlay.skipWorktree(),
                 "tracked copied paths are --skip-worktree'd");
     }
 

scripts/probe-member-credentials.sh

@@ -22,8 +22,34 @@
 #     A prefix of a short secret is most of the secret, and it would end up pasted into a ticket. The
 #     hash answers every question the prefix was for — is it set, is it the same value as over there,
 #     is it the CB-592 sentinel — and answers none of the ones it should not.
-#   * It never writes anywhere, never contacts the network, and never touches secrets.sh, which is
-#     the operator's file.
+#   * It never writes anywhere, never contacts the network except the daemon's own REST port (see
+#     below), and never touches secrets.sh, which is the operator's file.
+#
+# WHERE THE NAME LIST COMES FROM (fleetd #111 / CB-608)
+#
+# Earlier versions of this script carried their own hardcoded NAMES array, recorded by hand on
+# 2026-08-16. The live memberCredentials: policy in fleetd.yaml grew past that list, and this probe
+# never noticed — it kept checking the same 31 names, printed a clean-looking table, and exited 0.
+# A verification tool that silently under-reports the thing it verifies is worse than no tool at
+# all, because its "clean" output gets taken as proof rather than treated with the suspicion an
+# absent tool would get.
+#
+# The fix is the same one #114 used for the drifted tool catalogue: delete the hand-maintained copy
+# rather than update it. This script now fetches the policy's name list from the daemon itself, at
+# `GET /member-credentials` (dev.ltms.fleet.member.MemberCredentialPolicyView via FleetApp) — names
+# and counts only, the same way the daemon's own startup log line is computed, from the SAME class.
+# If fleetd adds a name to memberCredentials.known tomorrow, this probe checks it tomorrow too,
+# with no edit here required. There is no local fallback list. See fetch_policy() below for what
+# happens when the daemon cannot be reached — it is a hard failure, on purpose (see next section).
+#
+# WHY AN UNREACHABLE DAEMON IS A HARD FAILURE, NOT A DEGRADED RUN
+#
+# An empty (or short) name list passes every subset check trivially — a probe that checked zero
+# names would print "0 of 0 names are set" and look identical to a clean bill of health. That trap
+# has bitten this project twice in one week (see docs/memory — "silent defaults disable features"
+# and "a test on the seam does not prove the caller"). So the denominator is guarded explicitly:
+# this script refuses to proceed unless it got a policy with at least one known name, and it refuses
+# just as hard if the count it fetched does not match the count it is about to check.
 #
 # HOW TO RUN IT
 #
@@ -32,34 +58,22 @@
 #   2. For the comparison row, in your OWN shell — a lead, not a member:
 #        bash scripts/probe-member-credentials.sh --allow-outside-member
 #
+# Both readings need the daemon's REST port reachable (default http://127.0.0.1:8765; override with
+# FLEETD_HOST). That is normally true in every pane this script is meant to run in.
+#
 # The two outputs side by side are the finding: any name whose hash matches between them is a
 # credential the member holds in full.
 #
 set -uo pipefail
 
-# The names ${SHARED_ENV}/tools/secrets.sh exports, recorded on 2026-08-16 (issue #82). Names only —
-# this list contains no values and never should. If secrets.sh gains a name, this list goes stale and
-# the probe silently stops asking about it; that staleness is itself part of what #82's criterion 4
-# has to solve, so it is called out in the summary rather than hidden.
-NAMES=(
-  AI_GATEWAY_TOKEN BESZEL_ADMIN_EMAIL BESZEL_ADMIN_PASSWORD
-  BESZEL_HUB_URL BESZEL_KEY BESZEL_UNIVERSAL_TOKEN
-  BRAIN_MCP_TOKEN CF_ACCOUNT_ID CF_API_TOKEN
-  CF_USER_TOKEN CONFLUENCE_API_TOKEN CONFLUENCE_USERNAME
-  CONTEXT7_TOKEN GITEA_HOST GITLAB_OAUTH_CLIENT_SECRET
-  GITLAB_PERSONAL_ACCESS_TOKEN GRAFANA_ADMIN_PASSWORD GRAFANA_ADMIN_USER
-  HASS_TOKEN HW_PASSWORD HW_USER
-  LTMS_API_KEY MEMORY_MCP_TOKEN METRICS_PUSH_TOKEN
-  OPENCODE_AUTOMODE_MODEL TELEGRAM_BOT_TOKEN TELEGRAM_CHAT_ID
-  TS_API_KEY TS_AUTHKEY WORKER_GITEA_TOKEN
-  GITEA_ACCESS_TOKEN
-)
+FLEETD_HOST="${FLEETD_HOST:-http://127.0.0.1:8765}"
+POLICY_URL="${FLEETD_HOST%/}/member-credentials"
 
 allow_outside=0
 for arg in "$@"; do
   case "$arg" in
     --allow-outside-member) allow_outside=1 ;;
-    -h|--help) sed -n '2,40p' "$0"; exit 0 ;;
+    -h|--help) sed -n '2,60p' "$0"; exit 0 ;;
     *) echo "unknown argument: $arg" >&2; exit 2 ;;
   esac
 done
@@ -75,6 +89,107 @@ EOF
   exit 1
 fi
 
+# --- fetch the policy from the daemon (fleetd #111) — no local fallback, ever ------------------
+#
+# Prefer jq (a real JSON parser); fall back to python3 (present on every host this has run on so
+# far); if neither exists, fail loudly rather than guess at the JSON with grep/sed, which is exactly
+# the kind of "looks like it worked" degradation this ticket exists to remove.
+#
+# NOTE: jq's `//` alternative operator treats `false` AND `0` as "missing" and substitutes the
+# default — so `.present // empty` silently turns a real `"present": false` into an empty string
+# ("unknown"), not the false it actually is. Every extraction below reads its field directly
+# instead, so a genuine false/0 is reported as exactly that, not swallowed into "unknown".
+if ! command -v jq >/dev/null 2>&1 && ! command -v python3 >/dev/null 2>&1; then
+  echo "refusing to run: neither jq nor python3 is on PATH, and this probe will not guess at JSON" \
+       "with grep/sed. Install one of them, or run from a shell that has one." >&2
+  exit 1
+fi
+
+POLICY_JSON="$(curl -fsS --max-time 5 "$POLICY_URL" 2>/dev/null)"
+CURL_STATUS=$?
+if [ "$CURL_STATUS" -ne 0 ] || [ -z "$POLICY_JSON" ]; then
+  cat >&2 <<EOF
+refusing to run: could not fetch the memberCredentials policy from $POLICY_URL (curl exit $CURL_STATUS).
+
+This probe has NO built-in name list any more (fleetd #111) — it only checks what the live daemon
+reports, so an unreachable daemon means it cannot check anything at all. It will not fall back to a
+guessed or empty list, because an empty list would pass every check trivially and look clean.
+
+Fix: confirm fleetd is up (curl \${FLEETD_HOST:-http://127.0.0.1:8765}/healthz) and that
+FLEETD_HOST (if set) points at it, then re-run.
+EOF
+  exit 1
+fi
+
+# One parse pass: line 1 = present (true/false/null), line 2 = policy mode (possibly blank),
+# lines 3-5 = knownCount/allowedCount/blockedCount, remaining lines = the known[] names. A single
+# pass avoids re-parsing (and re-risking a truthiness bug) five separate times.
+if command -v jq >/dev/null 2>&1; then
+  mapfile -t _FIELDS < <(printf '%s' "$POLICY_JSON" | jq -r '
+      (.present | tostring),
+      (.policy // ""),
+      (.knownCount // 0 | tostring),
+      (.allowedCount // 0 | tostring),
+      (.blockedCount // 0 | tostring),
+      (.known[]? // empty)')
+else
+  mapfile -t _FIELDS < <(printf '%s' "$POLICY_JSON" | python3 - <<'PY'
+import json, sys
+data = json.load(sys.stdin)
+print(str(data.get("present")))
+print(data.get("policy") or "")
+print(data.get("knownCount") if data.get("knownCount") is not None else 0)
+print(data.get("allowedCount") if data.get("allowedCount") is not None else 0)
+print(data.get("blockedCount") if data.get("blockedCount") is not None else 0)
+for n in (data.get("known") or []):
+    print(n)
+PY
+  )
+fi
+
+PRESENT="${_FIELDS[0]:-null}"
+POLICY_MODE="${_FIELDS[1]:-}"
+KNOWN_COUNT_REPORTED="${_FIELDS[2]:-0}"
+ALLOWED_COUNT_REPORTED="${_FIELDS[3]:-0}"
+BLOCKED_COUNT_REPORTED="${_FIELDS[4]:-0}"
+NAMES=("${_FIELDS[@]:5}")
+
+# knownCount must be a plain non-negative integer for the arithmetic guard below — a malformed or
+# unparseable response must fail loudly, not be coerced into a number that happens to compare true.
+case "$KNOWN_COUNT_REPORTED" in
+  ''|*[!0-9]*)
+    echo "refusing to run: knownCount in the response ('$KNOWN_COUNT_REPORTED') is not a plain" \
+         "non-negative integer — the response could not be parsed as expected." >&2
+    exit 1
+    ;;
+esac
+
+# --- guard the denominator explicitly — never proceed on a zero/short count ---------------------
+#
+# This is the exact trap named in the ticket: an empty (or truncated) NAMES array passes every
+# subsequent "is it set" check vacuously and prints a table that LOOKS complete. So this is checked
+# before anything else runs, with a message that says why, not just that it failed.
+if [ "${#NAMES[@]}" -eq 0 ] || [ "$KNOWN_COUNT_REPORTED" -eq 0 ]; then
+  cat >&2 <<EOF
+refusing to run: the policy fetched from $POLICY_URL contains 0 known names (present=${PRESENT:-unknown}).
+
+Either memberCredentials: is absent/empty on the running daemon (nothing is protected — see fleetd's
+own startup warning), or the response could not be parsed. Either way, checking zero names would
+print a clean-looking table for a policy that protects nothing, or for a probe that read nothing.
+This is refused rather than reported as a pass.
+EOF
+  exit 1
+fi
+
+if [ "${#NAMES[@]}" -ne "$KNOWN_COUNT_REPORTED" ]; then
+  cat >&2 <<EOF
+refusing to run: the policy reports knownCount=$KNOWN_COUNT_REPORTED but the known[] array this probe
+parsed has ${#NAMES[@]} entries. That mismatch means the JSON was not parsed correctly, and this
+probe will not check a name list it cannot trust to be complete.
+EOF
+  exit 1
+fi
+
 # Prefer sha256sum (Linux), fall back to shasum (macOS). If neither exists, report presence and
 # length only — degraded, but never a value.
 hasher=""
@@ -95,12 +210,13 @@ else
   where="NOT a member — comparison reading only"
 fi
 
-echo "CB-596 credential probe"
+echo "CB-596 credential probe (fleetd #111: names sourced live from $POLICY_URL)"
 echo "reading from : $where"
 echo "shell        : ${SHELL:-unknown}"
 echo "hash         : ${hasher:-none available — lengths only}"
 # Only printed so the two readings can be told apart when they are pasted side by side.
 echo "host         : $(hostname 2>/dev/null || echo unknown)"
+echo "policy       : mode=${POLICY_MODE:-unknown} known=$KNOWN_COUNT_REPORTED allowed=${ALLOWED_COUNT_REPORTED:-?} blocked=${BLOCKED_COUNT_REPORTED:-?}"
 echo
 printf '%-30s %-7s %6s  %s\n' "NAME" "STATE" "LEN" "SHA256-12"
 printf '%-30s %-7s %6s  %s\n' "------------------------------" "-------" "------" "------------"
@@ -118,6 +234,7 @@ done
 
 echo
 echo "$set_count of ${#NAMES[@]} names are set in this shell."
+echo "policy contains $KNOWN_COUNT_REPORTED name(s); this run checked ${#NAMES[@]} — they match."
 echo
 cat <<'EOF'
 How to read this:
@@ -129,7 +246,8 @@ How to read this:
     most urgent thing on this page.
   * AI_GATEWAY_TOKEN matching is expected and correct, not a leak: fleetd.yaml names it in
     `tokenEnv:` for the local and gx profiles, so a member reaching the gateway is by design.
-  * A name that is set here but is NOT in the list above will not appear at all. The list was
-    recorded on 2026-08-16 and does not update itself. Anything added to secrets.sh since then is
-    invisible to this probe — which is the same gap issue #82 criterion 4 asks to close properly.
+  * The name list above is fetched live from the running daemon's memberCredentials: policy
+    (fleetd #111) — it is never hand-maintained here, so it cannot go stale the way the old
+    hardcoded list did. If the daemon's policy changes, the next run of this script reflects it
+    with no edit to this file.
 EOF