• Joined on 2026-03-08
ltms merged pull request fleet/fleetd#623 2026-09-22 06:38:30 +02:00
#608 replace MessageService timing sleeps
ltms pushed to main at fleet/fleetd 2026-09-22 06:31:02 +02:00
63eec8a0da Merge pull request 'fleetd #621: make the context-roll notice obey requireOperatorConfirm' (#622) from worker/621-b4520b-1 into main
cbe872b538 fleetd #621: make the context-roll notice obey requireOperatorConfirm
Compare 2 commits »
ltms closed issue fleet/fleetd#621 2026-09-22 06:31:02 +02:00
contextNotice hardcodes "ask the operator" and ignores requireOperatorConfirm, so the knob cannot actually stop the asking
ltms merged pull request fleet/fleetd#622 2026-09-22 06:31:02 +02:00
fleetd #621: make the context-roll notice obey requireOperatorConfirm
ltms pushed to main at fleet/fleetd 2026-09-22 06:24:12 +02:00
7d9a807243 handover skill: the rollover bootstrap is proven, and requireOperatorConfirm is per-host
ltms commented on issue fleet/fleetd#612 2026-09-22 06:22:38 +02:00
Fleetd.main's injected wirings are unpinned: 16 call sites go inert with a fully green suite

The owed check on FleetMcpAuthzTest — done. It will not go silently blind.

Architect 1 flagged this and did not chase it, and my decision comment listed it as "one grep before C is…

ltms commented on issue fleet/fleetd#612 2026-09-22 06:20:04 +02:00
Fleetd.main's injected wirings are unpinned: 16 call sites go inert with a fully green suite

Lead decision on the re-sequencing — two architects, they disagreed, I decided

I asked two architects the same question, blind to each other. They agree on most of it and disagree on one…

ltms opened issue fleet/fleetd#621 2026-09-22 06:01:58 +02:00
contextNotice hardcodes "ask the operator" and ignores requireOperatorConfirm, so the knob cannot actually stop the asking
ltms commented on issue fleet/fleetd#612 2026-09-22 05:59:17 +02:00
Fleetd.main's injected wirings are unpinned: 16 call sites go inert with a fully green suite

Operator decision: test files are the fleet's call. The handover's gate on them is lifted.

I asked the operator whether Unit A may open the three source-text test files beyond the three Unit D…

ltms commented on issue fleet/fleetd#612 2026-09-22 05:56:43 +02:00
Fleetd.main's injected wirings are unpinned: 16 call sites go inert with a fully green suite

Unit A is built and it turns the build red. The A→B→C→D order cannot work as designed.

Unit A landed as PR #620. I ran the gate the worker could not run, and it fails. This is not a worker…

ltms commented on issue fleet/fleetd#618 2026-09-22 05:54:02 +02:00
Measured: CLAUDE_CODE_AUTO_COMPACT_WINDOW beats --autocompact, so fleetd.yaml's comment is backwards

Follow-ups 1, 2 and 4 are done. Item 3 stays open, and it is a real decision.

1. fleetd/fleetd.yaml comments — done (lead, this host). The four comments on lines 25, 74, 110 and 124…

ltms pushed to main at fleet/fleetd 2026-09-22 05:53:37 +02:00
8915e40c7d Merge pull request 'fleetd #618: state the measured auto-compact precedence' (#619) from worker/618-b83894-2 into main
6cb31a10e4 fleetd #618: fix the third stale spot the brief missed
8368a274a0 fleetd #618: state the measured auto-compact precedence, not 'unverified'
Compare 3 commits »
ltms merged pull request fleet/fleetd#619 2026-09-22 05:53:36 +02:00
fleetd #618: state the measured auto-compact precedence
ltms commented on issue fleet/fleetd#618 2026-09-22 05:48:16 +02:00
Measured: CLAUDE_CODE_AUTO_COMPACT_WINDOW beats --autocompact, so fleetd.yaml's comment is backwards

Correction to the PR #619 brief: there is a THIRD stale spot, and my brief missed it

This is a correction to the brief, not a new unit. The brief named two places that still claim the…

ltms opened issue fleet/fleetd#618 2026-09-22 05:41:20 +02:00
Measured: CLAUDE_CODE_AUTO_COMPACT_WINDOW beats --autocompact, so fleetd.yaml's comment is backwards
ltms commented on issue fleet/fleetd#612 2026-09-22 05:23:25 +02:00
Fleetd.main's injected wirings are unpinned: 16 call sites go inert with a fully green suite

Architect proposal for the fix shape, with a delegation split

An architect worked this and settled on a shape. Recording it in full so implementation can start from it. I have not yet…

ltms commented on pull request fleet/fleetd#601 2026-09-22 05:23:06 +02:00
CB-617: pass auto-compact window to leads

Merged. The start-up hazard is gone — measured with the same probe that refused before.

The guard now warns instead of throwing, so this no longer has to wait for every host's config to be…

ltms pushed to main at fleet/fleetd 2026-09-22 05:22:57 +02:00
17127efb88 Merge #601: pass auto-compact window to leads; warn instead of refusing on a conflict (CB-617)
be6c45ff78 CB-617 review: warn instead of refuse on conflicting autoCompactWindow
e99cb70a8b CB-617: pass auto-compact window to leads
Compare 3 commits »
ltms merged pull request fleet/fleetd#601 2026-09-22 05:22:57 +02:00
CB-617: pass auto-compact window to leads
ltms closed issue fleet/fleetd#615 2026-09-22 05:21:34 +02:00
A herdr throw during a lead roll leaves status(token) reporting IN_PROGRESS forever