chore: remove broken coily-trailer prepare-commit-msg hook #59

New issue

Closed

opened 2026-06-06 05:57:21 +00:00 by coilysiren · 2 comments

coilysiren commented

2026-06-06 05:57:21 +00:00

Owner

Remove the broken coily-trailer prepare-commit-msg hook (hard-fails on coily v2.53.0, blocks all commits). Fleet decision: coilysiren/inbox#67.

Remove the broken `coily-trailer` prepare-commit-msg hook (hard-fails on coily v2.53.0, blocks all commits). Fleet decision: https://forgejo.coilysiren.me/coilysiren/inbox/issues/67.

coilysiren added the

label

2026-06-17 08:39:48 +00:00

coilysiren referenced this issue from coilyco-flight-deck/ward

2026-07-02 02:26:14 +00:00

warded dispatch: transient issue-resolve failure surfaces as a bare 'exit status 3' with no cause and no retry #497

coilyco-ops commented

2026-07-02 02:34:30 +00:00

Collaborator

🔒 Reserved by ward agent --driver claude — container engineer-claude-website-59 on host kais-macbook-pro-2.local is carrying this issue (reserved 2026-07-02T02:34:27Z). Concurrent ward agent runs are blocked until it finishes or the reservation goes stale (2h0m0s TTL); --force overrides.

— Claude (she/her), via ward agent

🔒 Reserved by `ward agent --driver claude` — container `engineer-claude-website-59` on host `kais-macbook-pro-2.local` is carrying this issue (reserved 2026-07-02T02:34:27Z). Concurrent `ward agent` runs are blocked until it finishes or the reservation goes stale (2h0m0s TTL); `--force` overrides.  — Claude (she/her), via `ward agent`

coilysiren referenced this issue from a commit

2026-07-02 02:40:39 +00:00

chore: confirm coily-trailer prepare-commit-msg hook already removed

coilysiren closed this issue

2026-07-02 02:40:39 +00:00

coilysiren commented

2026-07-02 02:41:29 +00:00

Author

Owner

WARD-OUTCOME: done - verified the coily-trailer prepare-commit-msg hook was already removed and landed a documenting close on main.

Honest retrospective: this one was a no-op in disguise. The task read like a real chore, but the first thing the git history told me was that the broken coily-trailer hook block had already been stripped from .pre-commit-config.yaml back on 2026-06-04 (commit 620dbfc), a full two days before this fleet issue was even filed. That commit was already an ancestor of main, and a full tree scan turned up zero residue - no coily-trailer, no trailer-hook, no stray prepare-commit-msg stage anywhere.

What fought back, mildly: the temptation to invent work. There was nothing to delete, so the honest move was an empty documenting commit rather than a fabricated change. The only real friction was the closes-issue hook rejecting the bare closes #59 trailer - it demands the full Forgejo URL form to avoid silently closing the wrong tracker on the GitHub mirror, which is a sharp little guardrail I hadn't front-loaded. Swapped to the URL form and it sailed through.

Confidence is high: HEAD matches origin/main, all pre-commit hooks passed, and the issue flipped to closed on push. Rough edge worth noting for the fleet: this is the classic race where a per-repo fleet issue lands after a repo already self-healed, so several sibling repos may also close as no-ops. No follow-up needed here.

WARD-OUTCOME: done - verified the coily-trailer prepare-commit-msg hook was already removed and landed a documenting close on main. Honest retrospective: this one was a no-op in disguise. The task read like a real chore, but the first thing the git history told me was that the broken `coily-trailer` hook block had already been stripped from `.pre-commit-config.yaml` back on 2026-06-04 (commit 620dbfc), a full two days *before* this fleet issue was even filed. That commit was already an ancestor of main, and a full tree scan turned up zero residue - no `coily-trailer`, no `trailer-hook`, no stray `prepare-commit-msg` stage anywhere. What fought back, mildly: the temptation to invent work. There was nothing to delete, so the honest move was an empty documenting commit rather than a fabricated change. The only real friction was the `closes-issue` hook rejecting the bare `closes #59` trailer - it demands the full Forgejo URL form to avoid silently closing the wrong tracker on the GitHub mirror, which is a sharp little guardrail I hadn't front-loaded. Swapped to the URL form and it sailed through. Confidence is high: HEAD matches origin/main, all pre-commit hooks passed, and the issue flipped to closed on push. Rough edge worth noting for the fleet: this is the classic race where a per-repo fleet issue lands after a repo already self-healed, so several sibling repos may also close as no-ops. No follow-up needed here.