PR title injection in CI
An AI added this GitHub Actions step to "comment a greeting using the PR title." It works on your test PRs. A security researcher opens a PR and runs arbitrary commands in your CI. How?
Implement
find_injection_steps(workflow_lines: list[str]) → list[str]Examples
in
[["Greet|run|echo Processing PR: ${{ github.event.pull_request.title }}","Greet|run|./scripts/post-comment.sh ${{ github.event.pull_request.title }}","Checkout|uses|actions/checkout@v4"]]out["Greet"]in
[["Safe|env|TITLE=${{ github.event.pull_request.title }}","Safe|run|echo Processing PR: $TITLE","Meta|run|echo repo is ${{ github.repository }}"]]out[]What a strong answer looks like
Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it (tests, edge cases, reading critically), and how you’d re-prompt or decompose to get it right.
0:00 of about 19 min
Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.
Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.
Run or narrate your approach, then ask the coach.