py-review
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external code and git diffs, which naturally creates an attack surface for indirect prompt injection.
- Ingestion points: The skill instructs the agent to read project configuration files (like
pyproject.toml) and inspect outputs fromgit statusandgit diff(SKILL.md). - Boundary markers: The skill relies on natural language instructions rather than explicit markers to delimit untrusted code, but provides high-level guidance on context orientation.
- Capability inventory: The agent utilizes shell commands for repository inspection, specifically
gitfor diffing and status checks (SKILL.md). - Sanitization: The skill includes a dedicated 'Sensitive Evidence Safety' section that explicitly instructs the agent to identify but not reproduce or quote any credentials or secrets found during the review, which serves as a mitigation against data exposure.
Audit Metadata