pr-fixer
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes PR review comments which represent an external, untrusted data surface. \n
- Ingestion points: The workflow involves reading all PR comments using the
gh apicommand to list comments (SKILL.md). \n - Boundary markers: There are no explicit markers or instructions to treat the content of comments as potentially untrusted data. \n
- Capability inventory: The agent has the capability to write files, run tests, and perform git commits based on the content of these comments. \n
- Sanitization: The skill lacks explicit sanitization or validation logic for the text within the comments before implementation.
Audit Metadata