agent-result-gate
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The
scripts/inspect-artifacts.pyfile uses__import__('sys')to access standard error reporting. This dynamic loading is restricted to a core Python library and does not introduce arbitrary execution vulnerabilities. - [INDIRECT_PROMPT_INJECTION]: This skill processes untrusted artifacts, establishing an attack surface for indirect prompt injection. Ingestion points: Files in the directory analyzed by
scripts/inspect-artifacts.py. Boundary markers: Strict safety criteria defined inreferences/inspection-rules.md. Capability inventory: Local file-read and execution of provided safety scripts. Sanitization: Automated rejection of symlinks and executables withinscripts/inspect-artifacts.py, and multi-check validation inscripts/validate-result-gate-report.py.
Audit Metadata