error-recovery
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a protocol for handling runtime error messages, which constitutes an attack surface (SKILL.md). Ingestion points: Runtime error messages from API failures, tool crashes, and environment issues. Boundary markers: The suggested error logging format does not utilize delimiters or specific instructions to ignore embedded content. Capability inventory: The instructions suggest the use of shell tools like 'lsof', process termination, and automated dependency installation. Sanitization: No input validation or sanitization of error content is mentioned.
- [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The documentation includes instructions for installing additional skills from the vendor's own repository using 'npx skills add jwilger/agent-skills'.
Audit Metadata