flux-image

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides links to documentation and installation guides hosted on the official GitHub repository for the inference-sh organization. These references are for the prerequisite belt CLI tool.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest user-provided prompts and image URLs for conversion into images.
  • Ingestion points: Prompt and image URL inputs specified in the belt app run command templates in SKILL.md.
  • Boundary markers: Inputs are encapsulated within a JSON object passed to the --input flag.
  • Capability inventory: The skill's execution environment is restricted to the belt command via the allowed-tools: Bash(belt *) configuration.
  • Sanitization: Input validation and sanitization are delegated to the belt CLI tool and the backend inference service.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 08:14 PM
Security Audit — agent-trust-hub — flux-image