comfyui-skill-openclaw

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses subprocess calls to interact with the 'comfyui-skill' CLI and perform system operations like 'git clone' and 'pip install' when managing ComfyUI custom nodes (found in ui/app.py and ui/dependency_installer.py).
  • [EXTERNAL_DOWNLOADS]: Fetches pre-built frontend assets from GitHub releases and allows downloading third-party ComfyUI node code from GitHub repositories (found in scripts/shared/frontend_update.py and ui/dependency_installer.py).
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests ComfyUI workflow JSON files which represent an ingestion surface for untrusted data, though the skill performs normalization and uses a defined schema to map parameters (found in ui/workflow_import.py).
  • [DYNAMIC_EXECUTION]: Uses os.execv to restart the FastAPI application process following a system update, replacing the current process with a new instance using the current python executable (found in scripts/shared/updater.py).
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 12:33 PM
Security Audit — agent-trust-hub — comfyui-skill-openclaw