Compare commits

..

3 commits

Author SHA1 Message Date
8ecfb4ec5b Merge pull request 'chore(labels): the review panel is glm, claude and kimi' (#278) from lead/panel-three-reviewers into main
All checks were successful
CI / release-exercise (push) Has been skipped
CI / self-guards (push) Successful in 14s
CI / action-exercise (push) Successful in 11s
CI / docs-sync-exercise (push) Successful in 11s
release / release (push) Successful in 12s
CI / test (push) Successful in 7m1s
Reviewed-on: #278
2026-10-07 10:40:35 +00:00
claude-lead-andresmgsl
d5df242522 docs(contributing): the roster is the three reviewers; builders listed apart
All checks were successful
CI / release-exercise (pull_request) Successful in 9s
CI / self-guards (pull_request) Successful in 6s
CI / action-exercise (pull_request) Successful in 4s
CI / docs-sync-exercise (pull_request) Successful in 5s
Refs guard / refs-not-closing (pull_request) Successful in 4s
labels / labels (pull_request) Successful in 7s
CI / test (pull_request) Successful in 3m46s
The previous commit moved panel= to glm, claude and kimi, and
test/labels.test.sh held the roster table to the old four, so the drift
check went red as it should.

- Roster table: codex's "build + review" row removed; the roster is the
  identities panel= and triage-actors= name, as roster_from_conf reads them.
- The panel paragraph no longer rests on "the only builder is itself on the
  panel and recuses": builders are off the panel, so a builder's PR needs all
  three, and only a panel identity's own PR drops to two. The rule
  (panel-minus-author) is unchanged.
- A new ### Builders section names codex and codex-2 and why they are off
  the panel. It sits under its own heading, which is where roster_from_doc
  stops reading.

Local: bash test/run.sh -> test files 32 passed, 0 failed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018EuKA4GdpAJMhBtPUj7phs
2026-09-30 20:44:22 +00:00
59b236ad49 chore(labels): the review panel is glm, claude and kimi
Some checks failed
CI / release-exercise (pull_request) Successful in 8s
CI / self-guards (pull_request) Successful in 4s
CI / action-exercise (pull_request) Successful in 3s
CI / docs-sync-exercise (pull_request) Successful in 3s
Refs guard / refs-not-closing (pull_request) Successful in 4s
labels / labels (pull_request) Successful in 6s
CI / test (pull_request) Failing after 3m59s
Operator decision, 2026-09-30: two builders (codex, codex-2) and three
reviewers (claude, kimi, glm). Builders do not review. Was: panel=codex-bot-andresmgsl glm-bot-andresmgsl claude-bot-andresmgsl kimi-bot-andresmgsl

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018EuKA4GdpAJMhBtPUj7phs
2026-09-30 18:35:45 +00:00
2 changed files with 19 additions and 8 deletions

2
.github/labels.conf vendored
View file

@ -1,4 +1,4 @@
panel=codex-bot-andresmgsl glm-bot-andresmgsl claude-bot-andresmgsl kimi-bot-andresmgsl panel=glm-bot-andresmgsl claude-bot-andresmgsl kimi-bot-andresmgsl
triage-actors=claude-bot-andresmgsl triage-actors=claude-bot-andresmgsl
scope:release-flow|C5DEF5|The reusable release workflow, decide, the doors scope:release-flow|C5DEF5|The reusable release workflow, decide, the doors
scope:guards|C5DEF5|changelog-armed / changelog-monotonic / drill-recorded scope:guards|C5DEF5|changelog-armed / changelog-monotonic / drill-recorded

View file

@ -41,7 +41,7 @@ only ceremony-specific facts such as the roster and code conventions.
### Roster ### Roster
Four identities share the work (org team `agents`), each living in its own Three identities review and triage (org team `agents`), each living in its own
[box](https://github.com/heavy-duty/box) — one box per credential, because [box](https://github.com/heavy-duty/box) — one box per credential, because
the box is the blast-radius boundary; roles are what a session is told, and the box is the blast-radius boundary; roles are what a session is told, and
[AGENTS.md](AGENTS.md) routes from there: [AGENTS.md](AGENTS.md) routes from there:
@ -49,17 +49,16 @@ the box is the blast-radius boundary; roles are what a session is told, and
| identity | box (rig tenant) | standing work | | identity | box (rig tenant) | standing work |
|---|---|---| |---|---|---|
| `claude-bot-andresmgsl` | `triage` (claude-box) | **triage** — the only door work issues come through; this identity mints work issues and nothing else writes them (#18's `triage-actors`) — and review. It does not build. | | `claude-bot-andresmgsl` | `triage` (claude-box) | **triage** — the only door work issues come through; this identity mints work issues and nothing else writes them (#18's `triage-actors`) — and review. It does not build. |
| `codex-bot-andresmgsl` | codex-box | build + review |
| `glm-bot-andresmgsl` | glm-box | review | | `glm-bot-andresmgsl` | glm-box | review |
| `kimi-bot-andresmgsl` | kimi-box | review | | `kimi-bot-andresmgsl` | kimi-box | review |
**The review panel for any PR is every bench identity except its author** — **The review panel for any PR is every bench identity except its author** —
recusal by construction, enforced by the reconciler (#10): the required recusal by construction, enforced by the reconciler (#10): the required
verdicts are the panel minus the PR's author. On this roster that resolves verdicts are the panel minus the PR's author. Builders are not on the
to **three** cross-vendor approvals of the current head, because the only panel, so for every PR a builder writes that resolves to **three**
builder is itself on the panel and recuses from its own PRs; the rule is cross-vendor approvals of the current head; only a PR authored by a panel
panel-minus-author, and three is what it currently comes to, not a second identity drops to two. The rule is panel-minus-author, and three is what it
rule. Builders and triage default to different models so the issue contract currently comes to, not a second rule. Builders and triage default to different models so the issue contract
is honestly exercised — a spec gap should surface as a question on the is honestly exercised — a spec gap should surface as a question on the
issue, not be silently filled by shared priors. Humans (`andres`) decide in issue, not be silently filled by shared priors. Humans (`andres`) decide in
proposals and merge; the roster is config, not doctrine — swapping a proposals and merge; the roster is config, not doctrine — swapping a
@ -81,6 +80,18 @@ ceremony's scopes are defined in [`.github/labels.conf`](.github/labels.conf)
[`.github/labeler.yml`](.github/labeler.yml). The conf is the set; no prose [`.github/labeler.yml`](.github/labeler.yml). The conf is the set; no prose
table repeats it (#104). table repeats it (#104).
### Builders
Builders write code and **do not review**, so they are deliberately absent
from the roster above and from `panel=` (operator decision, 2026-09-30).
Listing a builder on the panel makes every PR it did not author wait on a
verdict it will never give: the reconciler reads it as missing and lands
`state:addressing` + `blocker:unrequested` on PRs every real reviewer has
approved.
- `codex-bot-andresmgsl` — codex-box
- `codex-bot-2-andresmgsl` — codex-2 box
## Code conventions ## Code conventions
- Bash: `set -euo pipefail` in executables, `set -u` in test files (the test - Bash: `set -euo pipefail` in executables, `set -u` in test files (the test