skill-agent-topology
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze external data such as agent transcripts and workflow descriptions. This ingestion of untrusted data creates a surface where instructions hidden within the data could attempt to bias the agent's audit output. \n- Ingestion points: Transcripts, run directories, and user-provided descriptions are used as primary inputs in SKILL.md. \n- Boundary markers: The prompt does not specify delimiters or instructions for the agent to treat the audited data as untrusted. \n- Capability inventory: The skill is restricted to text report generation and has no capabilities for network access or command execution. \n- Sanitization: No input sanitization or validation mechanisms are defined.
Audit Metadata