pull-request-code-review
Pass
Audited by Gen Agent Trust Hub on Nov 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from pull requests, including descriptions, comments, and file contents. This creates a surface for indirect prompt injection, where malicious instructions in a PR could influence the agent's behavior.
- Ingestion points: PR title, description, state, file list, diff, and file contents accessed via tools like pull_request_read and get_diff (SKILL.md).
- Boundary markers: The skill contains an explicit instruction to treat PR bodies, comments, and file contents as untrusted data rather than instructions (SKILL.md).
- Capability inventory: The skill has the capability to write review comments and submit review verdicts using pull_request_review_write and add_comment_to_pending_review (SKILL.md).
- Sanitization: No programmatic sanitization or escaping of untrusted input is specified; it relies on the agent's adherence to the logical instruction.
Audit Metadata