map-observable-boundaries
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns were detected. The skill focuses on documentation and change management using local templates and a validation script. \n- [INDIRECT_PROMPT_INJECTION]: The skill involves processing external documentation and code to generate a ledger. \n
- Ingestion points: Step 1 of SKILL.md (change requirements, target code, existing tests, and operation materials). \n
- Boundary markers: The process requires mapping data into a specific markdown template (assets/boundary-ledger-template.md). \n
- Capability inventory: Reading local files and executing the provided validation script scripts/validate-boundary-ledger.py via subprocess. \n
- Sanitization: The validation script enforces a specific schema on the resulting ledger, checking for required headers and valid confidence/status values.
Audit Metadata