github-ops
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill functions by ingesting external data from GitHub issues and pull requests, which creates an inherent surface for indirect prompt injection attacks where untrusted data could contain malicious instructions.
- Ingestion points: The skill reads issue titles, bodies, comments, and PR metadata (SKILL.md).
- Boundary markers: No explicit boundary markers or 'ignore' instructions are provided to the agent to isolate external content.
- Capability inventory: The skill possesses write capabilities including
gh issue comment,gh issue edit,gh run rerun, andgh release create. - Sanitization: There is no explicit sanitization step defined for the content fetched from the GitHub API before processing.
Audit Metadata