github

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly instructs the agent to fetch and display user-generated GitHub content (e.g., "When given a GitHub issue URL... run gh issue view" and gh pr view), so it will ingest untrusted public issues/PRs/comments that could contain indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs at runtime to extract and fetch GitHub issue/PR content (e.g., https://github.com/user/repo/issues/123) via the gh CLI and use that issue/PR data as the agent's context, which directly controls prompts/instructions.