Skip to content

ci: pin maintenance workflow actions - #222

Merged
pierreraby merged 2 commits into
mainfrom
chore/pin-maintenance-actions
Nov 17, 2026
Merged

pierreraby merged 2 commits into
mainfrom
chore/pin-maintenance-actions

Conversation

@pierreraby

@pierreraby pierreraby commented Nov 17, 2026 •

Copy link
Copy Markdown
Collaborator

Summary

Pin 13 uses entries across the catalog-sync and memory-benchmark workflows to full commit SHAs. The nine distinct refs were resolved from their existing tags through the GitHub API, including annotated-tag dereferencing, and each pinned commit exposes an action.yml manifest. Original versions are retained in comments.

Only .github/workflows/model-metadata.yml and .github/workflows/memory-benchmark.yml change. A parsed-YAML comparison confirms that exactly 13 uses values changed; all other workflow values, inputs, scripts, permissions, triggers and job boundaries are identical to the base. No action version upgrade or runtime-provider change is included.

Validation

  • actionlint 1.7.12: PASS (ShellCheck disabled; no ShellCheck claim).
  • Unit typecheck and the entrypoint check against installed Pi 1.5.2: PASS.
  • Maintenance/release tests: 83/83 PASS.
  • Prettier on both workflow files and git diff --check: PASS.
  • Active YAML diagnostics: primary clean, no unpinned-uses findings on the changed files.

Known baseline / scope

This is independent of the catalog refresh following #220. The base still carries the older model snapshot, so the live catalog/pricing checks can fail on current upstream drift until the separate catalog refresh #223 lands. This pin-only change does not claim to resolve that drift.

Existing credential-persistence (artipacked) and ad-hoc-package warnings are unchanged. The guarded maintenance workflow_run workflow is also unchanged. No global all-scanners-clean verdict is claimed; no suppressions were added. No billable inference was performed.

Publication is for review/CI only, not an authorization to merge or release.

CI follow-up

At head c21e19732854836ff17d626f17a1f367485298e4, all executed checks pass except catalog-sync check (run 44052981565). Its logs confirm the known base drift: CLI 1.89.1 vs 1.89.2, plus missing Flint pricing on the 101-model API. Catalog PR #223 is green at 773c9e374371e9b43a4bba177d86afb3b57b8aaf.

Proposed integration order: review/merge #223 first when authorized, update this branch with main, then re-check the exact updated head. Do not bypass the failing check. No merge is being performed as part of this publication.

Resolve the existing nine action refs to immutable commit SHAs across thirteen uses entries. Preserve action versions, workflow inputs, permissions and triggers.
@github-actions

Copy link
Copy Markdown
Contributor

Dependency Review

✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.

OpenSSF Scorecard

PackageVersionScoreDetails
actions/actions/checkout 11d5960a326750d5838078e36cf38b85af677262 🟢 5.8
Details
CheckScoreReason
Code-Review🟢 10all changesets reviewed
Dangerous-Workflow🟢 10no dangerous workflow patterns detected
Binary-Artifacts🟢 10no binaries found in the repo
Maintained⚠️ 00 commit(s) and 0 issue activity found in the last 90 days -- score normalized to 0
Token-Permissions⚠️ 0detected GitHub workflow tokens with excessive permissions
CII-Best-Practices⚠️ 0no effort to earn an OpenSSF best practices badge detected
Pinned-Dependencies🟢 3dependency not pinned by hash detected -- score normalized to 3
Fuzzing⚠️ 0project is not fuzzed
License🟢 10license file detected
Packaging⚠️ -1packaging workflow not detected
Signed-Releases⚠️ -1no releases found
Security-Policy🟢 9security policy file detected
Branch-Protection🟢 5branch protection is not maximal on development and all release branches
SAST🟢 10SAST tool is run on all commits
actions/actions/setup-node 49933ea5288caeca8642d1e84afbd3f7d6820020 UnknownUnknown
actions/peter-evans/create-pull-request 3ffccacd46053766916f087ce41d627c0b8cf2d1 🟢 5.3
Details
CheckScoreReason
Maintained🟢 1012 commit(s) and 0 issue activity found in the last 90 days -- score normalized to 10
Binary-Artifacts🟢 10no binaries found in the repo
Packaging⚠️ -1packaging workflow not detected
Dangerous-Workflow🟢 10no dangerous workflow patterns detected
Code-Review🟢 3Found 2/6 approved changesets -- score normalized to 3
Token-Permissions⚠️ 0detected GitHub workflow tokens with excessive permissions
Pinned-Dependencies🟢 10all dependencies are pinned
CII-Best-Practices⚠️ 0no effort to earn an OpenSSF best practices badge detected
Fuzzing⚠️ 0project is not fuzzed
License🟢 10license file detected
Signed-Releases⚠️ -1no releases found
Branch-Protection⚠️ -1internal error: error during branchesHandler.setup: internal error: some github tokens can't read classic branch protection rules: https://github.com/ossf/scorecard-action/blob/main/docs/authentication/fine-grained-auth-token.md
Security-Policy⚠️ 0security policy file not detected
SAST⚠️ 0SAST tool is not run on all commits -- score normalized to 0

Scanned Files

  • .github/workflows/model-metadata.yml

@github-actions

github-actions Bot commented Nov 17, 2026 •

Copy link
Copy Markdown
Contributor

Memory benchmark

Compared base c995c01 with PR head ba872aa on the same GitHub-hosted darwin runner. Lower values are better.

Metric Base PR PR − Base Change
Stable RSS 107.3 MiB ± 0.6 MiB 107.4 MiB ± 0.6 MiB -0.0 MiB ± 0.5 MiB -0.0%
Physical footprint 60.9 MiB ± 0.4 MiB 60.9 MiB ± 0.4 MiB -0.1 MiB ± 0.6 MiB -0.2%
Physical peak 71.7 MiB ± 0.4 MiB 72.4 MiB ± 0.6 MiB +0.0 MiB ± 1.0 MiB +0.0%

Interpretation: No metric shows a clear Base-to-PR difference beyond its measured run-to-run variation.

Extension overhead above pi baseline

Metric Base overhead PR overhead Difference
Stable RSS +5.7 MiB ± 0.8 MiB +5.6 MiB ± 0.7 MiB -0.0 MiB ± 0.5 MiB
Physical footprint +3.7 MiB ± 0.7 MiB +3.7 MiB ± 0.4 MiB -0.1 MiB ± 0.6 MiB
Physical peak +3.8 MiB ± 1.2 MiB +4.4 MiB ± 1.3 MiB +0.0 MiB ± 1.0 MiB

Values are medians of 6 alternating, paired runs. The value after ± is the median absolute deviation (MAD). Each process was sampled 12 times after a 2500 ms warm-up.

Environment: pi 0.84.4, Bun 1.4.4, darwin arm64. RSS comes from ps; physical footprint and peak come from macOS footprint.

This is a comparative signal, not a pass/fail threshold. GitHub-hosted runner noise can affect absolute values.

@pierreraby

Copy link
Copy Markdown
Collaborator Author

Final maintainer verification at updated head ba872aa995e3f909d23808f1de295f0c4eed7a00: main now includes #223 (c995c01) and the net diff is still exactly the original 13 pins in two workflow files. Fresh actionlint, live catalog/pricing checks, real-host entrypoint typecheck, workflow formatting, and the complete Pi/OMP-required mock suite pass locally. All checks for this updated head are now green, including catalog-sync run 44056253805; CI run 44056253766 and benchmark 44056253733 are green as well.

The earlier failure on c21e197 was the inherited catalog drift, now resolved by integrating reviewed main. Existing credential-persistence/ad-hoc-install advisories remain outside this pin-only diff; no suppressions were added.

The operator explicitly authorized merging this PR after the exact updated-head CI passed. No bypass, release, production update or billable inference is included.

@pierreraby
pierreraby merged commit 7f96417 into main Nov 17, 2026
15 checks passed
@pierreraby
pierreraby deleted the chore/pin-maintenance-actions branch November 17, 2026 09:34
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.

1 participant