Skip to content

Move the review workflow to code-review v0.3.0, which fetches the pull request itself - #18

Merged
polewskm merged 1 commit into
mainfrom
ci/code-review-v0-3-0
Oct 31, 2026
Merged

polewskm merged 1 commit into
mainfrom
ci/code-review-v0-3-0

Conversation

@polewskm

Copy link
Copy Markdown
Contributor

Moves the review workflow to NCodeGroup/code-review@v0.3.0 and removes its second checkout, the one of the pull request's head.

From v0.3.0 the engine fetches the exact commit itself, after it has decided to review, with git started without configuration, hooks or a credential helper. The workflow no longer checks out untrusted code in a privileged context, so CodeQL stops reporting it: the two alerts on this file ("checkout of untrusted code in a privileged context" and its TOCTOU variant) were accepted risks, dismissed by hand, and came back on every repository that adopted the bot.

Nothing else changes: the triggers, the code-review environment, the secrets, the permissions (the job token still has contents: read, which the engine uses for the fetch) and the default-branch checkout for the conventions. A /review after this merges is the first run of v0.3.0 here.

…l request itself

The engine fetches the pull request's head from v0.3.0 on, so the workflow drops its checkout of it, and CodeQL no longer reports the workflow.
@polewskm
polewskm deployed to code-review October 31, 2026 01:26 — with GitHub Actions Active

@ncode-review ncode-review Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

🤖 NCode Review at 143bde2: 1 new finding on the changed lines.

Comment thread .github/workflows/code-review.yml Resolved
@ncode-review

ncode-review Bot commented Oct 31, 2026

Copy link
Copy Markdown

🤖 Review complete

Verdict: 💬 Comment — The PR moves the review workflow to NCodeGroup/code-review v0.3.0, which fetches the pull request itself, so the workflow no longer checks out the PR head. It also updates the header comment to say this. The change is small and makes the design simpler: the pull_request_target job no longer puts untrusted head code on the runner, which lowers risk. The only open point is a pinning question that the file's documented convention already addresses.

4 reviewers run — 1 finding (run details)

Reviewer Model Findings Top
synthesis opus-5-5 1 🟡
conventions sonnet-5-5 0 🟢
hygiene haiku-5-5 0 🟢
rule-check sonnet-5-5 0 🟢
security sonnet-5-5 1 🟡

Skipped: api-compat, architecture, completeness, correctness, data-model, infrastructure, intent, performance, tests

Findings
  • 🟡 F1 .github/workflows/code-review.yml:59: Should the secret-bearing action be pinned to a commit SHA rather than to tag v0.3.0?

Trial: findings are posted as review threads and in the check run. $0.17 on Amazon Bedrock, 33s, at 143bde2. (how to interact)

@polewskm
polewskm marked this pull request as ready for review October 31, 2026 01:29
@polewskm
polewskm deployed to code-review October 31, 2026 01:29 — with GitHub Actions Active
@polewskm
polewskm merged commit 4e553bd into main Oct 31, 2026
11 checks passed
@polewskm
polewskm deleted the ci/code-review-v0-3-0 branch October 31, 2026 01:29

This branch was successfully deployed

1 active deployment
code-review — 143bde2f Deployed Oct 31, 2026 by polewskm via review #30
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