fleetd #743: an observer may now send to another observer

Three claims in the canonical block went false when the observer SEND grant
merged, and this file is the instruction surface the bridge ships:

  - the observer role definition said "never SEND"
  - invariant 3 listed send as "lead, architect, or collaborator"
  - the hand-opened-pane row said such a pane "cannot fleet_send back"

The grant is narrow. Authz permits an observer's SEND only when
CallerResolver.sendableObserverTarget() classifies the target as an observer
too, so a lead, a collaborator, an architect slot and a spawned member are
each refused. TASK_READ stays denied, so #705 remains closed.

Measured on the merged tree: mvn clean install exit 0, 2137 tests from 178
surefire files, 0 failures. Mutating away the grant's call site
(FleetMcp.java:474) kills exactly FleetMcpObserverSendDeliveryTest, so the
behaviour is pinned and not only the predicate.

CLAUDE.md

@@ -33,8 +33,9 @@ gate uses. A worker also carries its `sessionId`, `profile`, `worktree` and `bra
 carries the slot name it was bound to; a collaborator carries its registry name and its own
 `sessionId`, and **no `leader` key** — a collaborator is a named peer, not a primary. An **observer**
 carries only its own `sessionId`: a pane the daemon could not place as any of the above, authorized
-to `READ`/`METRICS` and to `REPLY`/`ASK` on its own pane and nothing more — never `SEND`, never a
-ticket. Don't infer what you can ask.
+to `READ`/`METRICS`, to `REPLY`/`ASK` on its own pane, and to `SEND` only to a target that resolves
+as an observer too — never to a lead, a collaborator, or a spawned member, and never a ticket.
+Don't infer what you can ask.
 
 Only if that call is unavailable, fall back to these — each is one-way, so keep reading until one
 fires: the reply charter in your system prompt (*"You are a spawned member in the
@@ -62,9 +63,10 @@ and the sender silently receives nothing. Fail toward the recoverable error.
 2. **The bridge is the only channel.** Text you print in your terminal reaches nobody — the other
    side cannot see your screen. An answer that isn't in a `fleet_*` call is silently discarded.
 3. **Identity comes from the connection, never an argument.** Workers never pass a target; you
-   cannot act as another session. Spawn/stop/drain are lead-only; **send is lead, architect, or
-   collaborator** — and a collaborator may send only to a lead or another collaborator, never to a
-   spawned member's terminal; reply/ask are only-as-itself — any peer may answer for its own pane,
+   cannot act as another session. Spawn/stop/drain are lead-only; **send is lead, architect,
+   collaborator, or observer** — and a collaborator may send only to a lead or another collaborator,
+   never to a spawned member's terminal, while an observer may send only to another observer pane;
+   reply/ask are only-as-itself — any peer may answer for its own pane,
    and for no other. A call outside your role is refused, not queued.
 4. **Delivery is status-gated: one message per turn.** Don't busy-poll a peer's terminal and don't
    re-send because a call looks slow — the bridge delivers when the peer is `idle`, `blocked` or
@@ -178,7 +180,7 @@ you decide.
 | Message a **peer lead** on this host | `fleet_send{sessionId: <their terminal>, content}` — `fleet_list` → `leads` reports it. Coordination only, **never** a task |
 | Message a **peer lead** on another daemon or host | `fleet_send{coordId: <their coord-id>, content}` — needs a `coordinator:` block; your own coord-id is in `fleet_list`. Coordination only, **never** a task |
 | Message a **collaborator** on this host | `fleet_send{sessionId: <their terminal>, content}` — `fleet_list` reports a `collaborators` array, and each row carries that peer's `name` and the `sessionId` you send to. It is visible to you, to an architect and to another collaborator, never to a worker. Coordination only, **never** a task |
-| Message an **unconfigured pane** — a tab a person opened by hand | `fleet_send{sessionId: <their terminal>, content}` — it needs **no** `fleet.collaborators` entry and no restart, because a pane becomes deliverable the moment its agent connects the bridge MCP. **Neither `fleet_list` nor `ListAgents` lists these**, so read the terminal id by joining `herdr tab list` to `GET /agents` on `tab_id`. Such a pane resolves as an `observer`: it can answer you with `fleet_reply`, and it cannot `fleet_send` back. Coordination only, **never** a task |
+| Message an **unconfigured pane** — a tab a person opened by hand | `fleet_send{sessionId: <their terminal>, content}` — it needs **no** `fleet.collaborators` entry and no restart, because a pane becomes deliverable the moment its agent connects the bridge MCP. **Neither `fleet_list` nor `ListAgents` lists these**, so read the terminal id by joining `herdr tab list` to `GET /agents` on `tab_id`. Such a pane resolves as an `observer`: it can answer you with `fleet_reply`, and it can `fleet_send` to another observer pane, but never to you. Coordination only, **never** a task |
 | Answer a peer lead that messaged you | `fleet_send{coordId}` — or `{sessionId}` if they are on this host. **Not** `fleet_reply`: it has no peer route and the publish is refused |
 | Read your own held lead-to-lead mail (no ack) | `fleet_poll{coordId: <your own coord-id, from fleet_list's coordinator.selfId>}` — primary-only; never acks, so `fleet_list`'s `held[]` still shows it after. `fleet_list`'s `held[]` gives only a truncated preview — this is the only way to read the full body |
 | Collect a held reply | `fleet_poll{target}` · then `fleet_ack{target, msgId}` |