Skills
skills/agentspace-so/runcomfy-agent-skills/video-edit/Gen Agent Trust Hub

video-edit

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes the runcomfy CLI to interact with remote video processing models as described in SKILL.md.- [EXTERNAL_DOWNLOADS]: Functional use requires the installation of the @runcomfy/cli package from a public registry.- [CREDENTIALS_UNSAFE]: The skill manages authentication tokens stored locally at ~/.config/runcomfy/token.json or passed via environment variables.- [PROMPT_INJECTION]: 1. Ingestion points: User-provided text prompts and external media URLs (video/image) in SKILL.md. 2. Boundary markers: Present; input is encapsulated in a JSON structure within CLI arguments. 3. Capability inventory: Execution of shell commands via the runcomfy CLI in SKILL.md. 4. Sanitization: Absent; the skill relies on the downstream API provider to filter malicious or adversarial prompt content.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 07:50 AM