slice
Pass
Audited by Gen Agent Trust Hub on Jun 30, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill ingests untrusted data from external GitHub issues via
gh issue viewand local repository documentation to draft new issues. This creates a surface for indirect prompt injection where malicious content in an issue could attempt to influence the agent's behavior. - Ingestion points: Output of
gh issue viewand local files such asDOMAIN-AWARENESS.mdandCONTEXT.md. - Boundary markers: No explicit delimiters or boundary markers are used to separate untrusted external content from the agent's internal instructions.
- Capability inventory: The skill is authorized to perform write operations to the repository using
gh issue create. - Sanitization: While no automated sanitization is defined, the skill implements a mandatory human-in-the-loop (HITL) step requiring the user to approve the breakdown before any issues are created.
Audit Metadata