prepare-pr

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and reads user-generated GitHub PR content (via gh pr view and git fetch origin pull/<PR>/head) and consumes .local/review.md review findings as mandatory input, which the agent uses to decide and perform fixes, commits, and pushes—so untrusted third-party content can materially influence its actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill fetches and checks out remote PR code at runtime (git fetch origin pull//head:pr-) and adds/uses the PR repository URL ($head_repo_url.git, e.g. https://github.com/owner/repo.git or git@github.com:owner/repo.git), then runs build/test (pnpm ...) which executes code from that remote repo.