comfyui-mcp
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes workflow metadata and notes that may originate from untrusted sources.
- Ingestion Points: Workflow notes accessed via the story_comfy tool.
- Capability Inventory: The agent can inspect, validate, and run ComfyUI workflows through a managed runtime.
- Boundary Markers: The instructions do not specify explicit delimiters but provide a high-level warning.
- Sanitization: The skill mandates that the agent treat this data as untrusted and requires explicit user approval for high-risk actions like node installation or model downloads.
Audit Metadata