evaluation-audit
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process and execute untrusted artifacts from the project environment, which could contain malicious instructions designed to influence the agent's behavior during the audit process.
- Ingestion points: Evaluation artifacts including runner/grader code, case sets, labels, and judge prompts (SKILL.md, references/health-checklist.md).
- Boundary markers: Absent; the instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when handling artifact content.
- Capability inventory: The agent is instructed to execute local code ('run the eval end to end') and read diverse project files (SKILL.md).
- Sanitization: Absent; there is no requirement for the agent to sanitize or validate the content of the artifacts before execution or inclusion in its reasoning context.
- [COMMAND_EXECUTION]: The skill requires the agent to execute code within the project's evaluation pipeline to verify functionality ('run the eval end to end on a handful of cases'). This is the primary intended function of the skill for performing technical audits.
Audit Metadata