aifollowup
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the local repository (unstaged changes and commits), which creates a potential surface for indirect prompt injection if the code being reviewed contains malicious instructions.
- Ingestion points: Workflow step 1 in
SKILL.mdrequires reading code from unstaged changes and recent commits. - Boundary markers: The instructions lack explicit delimiting or warnings to ignore commands that might be embedded in the code comments or content.
- Capability inventory: The skill uses file and git status read capabilities to provide context for its analysis. It does not contain instructions for network access, file writing, or command execution.
- Sanitization: No specific sanitization or filtering logic is defined for the ingested code content.
Audit Metadata