agent-security-audit
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is intended to process untrusted external data, specifically AI agent configurations, orchestration code, and RAG documents. This creates a surface for indirect prompt injection where the data being audited could contain instructions meant to deceive the auditing agent.
- Ingestion points: Reviewing
CLAUDE.md, MCP configurations, and agent orchestration code as specified in the description. - Boundary markers: The skill does not define specific delimiters or "ignore instructions" markers in the provided text.
- Capability inventory: The file contains no tools, scripts, or subprocess calls.
- Sanitization: No sanitization or validation logic is defined for the analyzed inputs.
- [NO_CODE]: The skill consists entirely of markdown instructions and YAML metadata. It does not include any scripts, executables, or automated tool definitions, which minimizes the direct risk to the host system.
Audit Metadata