ai-dev-loop
Pass
Audited by Gen Agent Trust Hub on Apr 22, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The files contain only documentation and instructions for a development process. No executable code or suspicious network activity was found.\n- [PROMPT_INJECTION]: The skill defines a workflow where agents process task files and linked PRDs, which represents an indirect prompt injection surface. This capability is essential for the development loop functionality. Standard platform-level safety measures are expected to handle external inputs.\n
- Ingestion points:
.agents/TASKS/*.mdand linked PRDs.\n - Boundary markers: Markdown headers serve as structural delimiters.\n
- Capability inventory: File system access, git operations, and interaction with development platforms.\n
- Sanitization: Not explicitly defined in the provided instructions.
Audit Metadata