analyze-agent-overlap
Pass
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE] (SAFE): No malicious patterns or code were detected. The skill is purely instructional markdown.
- [DATA_EXPOSURE] (SAFE): The skill accesses local agent directories to inventory files. No credentials or sensitive system paths are accessed.
- [INDIRECT_PROMPT_INJECTION] (SAFE): 1. Ingestion points: Reads configuration files from .github/, .claude/, and .codex/. 2. Boundary markers: Absent. 3. Capability inventory: Limited to text processing and reporting; no command execution or network capabilities. 4. Sanitization: Absent. The risk is safe due to the absence of exploitable tools.
Audit Metadata