Skip to content

ci: turn off CodeQL - #38

Merged
Birfy merged 1 commit into
ThinkFlowLab:mainfrom
Birfy:no-codeql
Oct 14, 2026
Merged

Birfy merged 1 commit into
ThinkFlowLab:mainfrom
Birfy:no-codeql

Conversation

@Birfy

@Birfy Birfy commented Oct 14, 2026

Copy link
Copy Markdown
Collaborator

Removes .github/workflows/codeql.yml (added in #29), at the maintainer's request.

  • No other references existed: no README badge, no docs, and no CodeQL-specific Dependabot entry (the github-actions ecosystem entry stays for the other workflows). tests-ok never depended on it.
  • Code scanning default setup is already not-configured, so no CodeQL analysis runs after this.
  • Existing code-scanning alerts are left as they are (not dismissed or deleted).
  • The Test workflow (ubuntu-latest · Node 24 · DSH 0.2.0-rc.2, package, tests-ok), Compatibility (nightly) and Release are unchanged.

@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 14, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-14T13:49:17.396173Z 40dd919 PR opened
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@Birfy
Birfy merged commit 253b2f9 into ThinkFlowLab:main Oct 14, 2026
3 checks passed
Birfy added a commit that referenced this pull request Oct 23, 2026
#814) (#38)

allPages ended at 20 pages of 100 without saying so, so a listing of 2000+ items was reported
as a complete total; it now returns whether the last allowed page came back full, and the
report prints "at least 2000 items, possibly truncated" for that repository with a note that
the counts are a lower bound.

skipped was filed under other and counted in the success-rate denominator, so 49 runs that
measured nothing held the 2026-10-22 rate down. It now has its own bucket, its own column and
stays out of the denominator with unknown.

The 2026-10-22 report is regenerated: 720/776 = 92.8% overall instead of 720/825 = 87.3%.
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