session-review
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from session logs and git history to extract behavior patterns and prepare handoffs.\n
- Ingestion points: Reads agent session logs and progress files from the local filesystem.\n
- Boundary markers: Not explicitly defined in the processing scripts.\n
- Capability inventory: Utilizes project management tools and standard file operations.\n
- Sanitization: Uses regex-based parsing to extract metrics and events from text logs.\n- [EXTERNAL_DOWNLOADS]: Fetches and runs the
@stevestomp/ohno-clitool via npx to retrieve project status and metrics.\n- [COMMAND_EXECUTION]: Runs local scripts to perform git analysis and process task data for reporting purposes.
Audit Metadata