trailsnap-commit
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The instructions require the agent to read and analyze CI pipeline logs via the GitHub CLI to perform root cause analysis and implement fixes (SKILL.md). This ingestion of external, untrusted log data without boundary markers represents a standard indirect prompt injection surface for development agents.- [SAFE]: The skill implements best practices for software development, such as Conventional Commits, mandatory Issue-PR linking, and CLA compliance.
Audit Metadata