Skip to content

docs: Record the Auto approval level, the requested-scope start, and the first-release npm-package step (#3, #5, #6) - #9

Merged
immakermatty merged 4 commits into
mainfrom
lazurio/steward-teams-run-on-auto
Sep 30, 2026
Merged

immakermatty merged 4 commits into
mainfrom
lazurio/steward-teams-run-on-auto

Conversation

@immakermatty

@immakermatty immakermatty commented Sep 30, 2026 •

Copy link
Copy Markdown

Status: ready for review. Docs only. Release 0.1.91-lazurio.1 is published, so merging no longer affects it.

Why

This PR records three decisions and facts from 2026-09-30 in the fork docs.

What changes

  • LAZURIO.md, docs/lazurio-github-intake.md and the release runbook say the operator sets all four bots to Auto after the import. Full and Custom stay desktop-only.
  • The intake example uses the requested scope. The organization scope, with its exclusions, is described as a later switch.
  • The runbook says to cancel the first npm-package.yml run before disabling the workflow, and records that this was done.

What does not change

There is no code change, and nothing lets a bot go above Auto.

Verification

  • The release contract test passes (6/6).
  • Release v0.1.91-lazurio.1:
    • it is immutable and latest, and its tag is e115bdda;
    • SHA256SUMS passes, and the attestation is signed by lazurio-release.yml@refs/heads/main;
    • release-evidence.json names upstream v0.1.91 (10786818).
  • npm-package.yml run 36685771083: build job steps through "Save the proven tarball" succeeded (artifact npm-package); publish to npm and update GitHub release were cancelled before starting. The release still has only its three assets.

Follow-ups

The pilot checks how Codex's sandbox treats git and gh in worktrees on Auto (#3).

🤖 Generated with Claude Code

The Organization Admin decided that unattended Steward teams run on the
Auto approval level. Imported bots start on Ask, so the operator sets all
four bots to Auto right after the import. Full and Custom stay desktop-only
as upstream, and the overlay adds no way past Auto. LAZURIO.md, the GitHub
intake doc and the release runbook say so instead of calling the level open.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…pe (#5)

The Organization Admin decided that the first enable does not use the
organization scope, whose first poll queues a review for every open ready
pull request. The example now shows the default requested scope, and the
organization scope with its exclusions is described as a later, deliberate
switch.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…n disable it (#6)

The first release, 0.1.91-lazurio.1, confirmed that the App-created tag
registers npm-package.yml and also starts it. The runbook now says to cancel
that run before disabling the workflow, and it records that this was done on
2026-09-30, before the run built anything. Later releases do not start it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@immakermatty immakermatty changed the title docs: Steward teams run on Auto (#3) docs: Record the Auto approval level, the requested-scope start, and the first-release npm-package step (#3, #5, #6) Sep 30, 2026

@agentrozjedemeai agentrozjedemeai left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Blocking factual correction in docs/operations/lazurio-fork-release.md:390-392: run 36685771083 was not cancelled before it built anything. The GitHub job list shows pnpm install --frozen-lockfile succeeded at 07:48:15Z, pnpm build:server && pnpm exec vite build && node scripts/build-npm-package.mjs succeeded at 07:48:55Z, the packed-tarball smoke succeeded at 07:49:00Z, and Save the proven tarball succeeded at 07:49:02Z. The run has an npm-package artifact. The publish to npm and update GitHub release jobs were cancelled with no steps, so the accurate statement is that cancellation happened after build/artifact upload but before those publish jobs. Please correct this sentence (and the corresponding issue #6 record if maintained) rather than recording the inaccurate operational evidence.

The other requested checks passed at exact head 4a20866: intake defaults to requested in server/github-intake.ts:87 and its test; the four-agent team package has no approval field, and server/package-import.ts:389-391 starts imported bots on Ask. Release v0.1.91-lazurio.1 exists with the archive, evidence JSON, and SHA256SUMS; npm-package.yml is disabled, while only the three lazurio-* workflows are active. Exact-head Fork CI passed all six jobs; review threads and pending review requests are zero. No merge.

Review of #9: run 36685771083 was not cancelled before it built anything. Its
build job had already built and saved the npm tarball as an Actions artifact;
only the npm publish and GitHub release update jobs were cancelled before
they started. The runbook now says so.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

@agentrozjedemeai agentrozjedemeai left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@immakermatty Re-review of exact head 41ed008: approved. Compared with 4a20866: only docs/operations/lazurio-fork-release.md changed, correcting the earlier factual blocker. In run 36685771083 the build, packed-tarball smoke, and "Save the proven tarball" steps succeeded; the npm-package artifact exists. The "publish to npm" and "update GitHub release" jobs were cancelled with no steps. Issue #6 comment and PR body now agree; the release still has three assets. Exact-head Fork CI: 6/6 jobs passed; zero review threads and no pending review requests. Local diff --check passed. No merge.

@immakermatty
immakermatty merged commit ca0e54f into main Sep 30, 2026
7 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants