engineering-pr-review
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted code diffs and repository files. * Ingestion points: Code diffs and repository content are read during context gathering in
SKILL.md. * Boundary markers: No explicit delimiters are used to separate code content from instructions. * Capability inventory: The skill can executepytestand linters. * Sanitization: Untrusted code content is not sanitized before processing. - [COMMAND_EXECUTION]: The skill directs the agent to run linters and test suites against the repository code.
- [EXTERNAL_DOWNLOADS]: The skill includes an instruction to install dependencies from the author's repository using the
npxcommand.
Audit Metadata