runcomfy-cli

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill documents potential indirect prompt injection vectors arising from third-party media URLs or web search results handled by remote models.
  • Ingestion points: Input JSON structures or prompt batches (prompts.txt) supplied to the runcomfy run command.
  • Boundary markers: The documentation advises wrapping inputs in a structured JSON body inside single quotes (--input '<JSON>') to establish clear boundaries.
  • Capability inventory: Uses Bash(runcomfy *) to execute model endpoints, list schemas, and manage deployments or datasets.
  • Sanitization: The CLI avoids shell expansion by transmitting the structured JSON payload directly to the API over HTTPS, reducing shell-injection risks.
  • [SAFE]: The installation instructions utilize the verified npm registry for the official vendor package @runcomfy/cli and explicitly advise against piping unverified remote installation scripts into the shell.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 04:16 AM
Security Audit — agent-trust-hub — runcomfy-cli