scan-backend-patterns
Pass
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: No malicious behavior, data exfiltration, or unauthorized command execution was found. The skill's operations are confined to the local file system for documentation purposes.- [NO_CODE]: The skill is composed strictly of markdown-based instructions for the AI agent and does not contain or download external scripts or binaries.- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it processes untrusted source code. 1. Ingestion points: Codebase source and configuration files. 2. Boundary markers: None specified. 3. Capability inventory: Local file read/write only. 4. Sanitization: None specified. The risk is assessed as safe given that the tool's impact is limited to documentation generation.
Audit Metadata