skill-trace-evaluation
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes content from external, untrusted agent skills, creating a surface for indirect prompt injection attacks where the analyzed content could attempt to influence the evaluation results.
- Ingestion points: The skill reads the target skill's
SKILL.mdfile and the contents of itsreferences/andexamples/directories via thetrace_evaluate.pyscript and direct AI reading. - Boundary markers: The instructions do not specify the use of clear delimiters or 'ignore instructions' warnings when the AI processes the content of target skills.
- Capability inventory: The skill performs file system reads on the target directory and writes report files (
trace-report.mdandevaluation-report.html) to the local file system. - Sanitization: The ingested text is not sanitized or filtered before being processed by the AI for semantic evaluation.
Audit Metadata