video-inpainting

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the @runcomfy/cli package from the official npm registry. This tool is the primary mechanism for the skill's functionality and is managed by the service provider described in the documentation.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes external media content.
  • Ingestion points: Untrusted data enters the agent context via the video_url parameter in the runcomfy run command.
  • Boundary markers: There are no technical delimiters used for the input URL; however, the skill documentation includes explicit warnings for the agent to only process user-provided URLs.
  • Capability inventory: The skill uses the runcomfy CLI which performs network requests to external APIs and writes files to the local system.
  • Sanitization: No automated sanitization or filtering of the external video content or its metadata is performed prior to processing.
  • [COMMAND_EXECUTION]: The skill executes the runcomfy CLI tool. The frontmatter configuration limits the scope of shell access to this specific command, which follows the principle of least privilege.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 10:56 AM
Security Audit — agent-trust-hub — video-inpainting