ai-content-pipeline

Pass

Audited by Gen Agent Trust Hub on May 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill serves as documentation for the inference.sh platform, guiding users on how to orchestrate AI models using the vendor's CLI tool. All referenced domains and repositories are vendor-owned.
  • [COMMAND_EXECUTION]: The skill contains multiple examples of using the Bash tool to run the 'belt' CLI for various AI tasks. This is consistent with the skill's stated purpose.
  • [EXTERNAL_DOWNLOADS]: The skill references the vendor's official GitHub repository (inference-sh) for additional components and setup instructions.
  • [PROMPT_INJECTION]: The multi-step pipelines ingest data from previous AI steps (e.g., text for voiceover), which constitutes an indirect prompt injection surface.
  • Ingestion points: Data enters the context from intermediate JSON files as seen in SKILL.md.
  • Boundary markers: No delimiters or explicit warnings are used in the command templates.
  • Capability inventory: The skill uses the Bash tool to execute belt commands in SKILL.md.
  • Sanitization: No sanitization is implemented for the data passed between stages.
Audit Metadata
Risk Level
SAFE
Analyzed
May 17, 2026, 08:02 PM
Security Audit — agent-trust-hub — ai-content-pipeline