happyhorse

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the belt-sh/cli utility using npx and references configuration guidelines hosted on the official inference.sh GitHub repository. These resources are part of the platform's standard setup.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides an interface for processing external media files (images and videos) as part of the video generation and editing workflow, which creates a surface for instructions hidden within data to influence agent behavior.
  • Ingestion points: Data enters the context via the first_frame, reference_images, and video parameters used in the belt app run commands (SKILL.md).
  • Boundary markers: The instructions do not define specific delimiters or instructions to ignore potential commands embedded in provided media.
  • Capability inventory: The skill is authorized to use the Bash tool to execute belt commands, which can interact with local and remote resources (SKILL.md).
  • Sanitization: No evidence of input validation or content filtering for the processed media files is present in the skill definition.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 01:13 PM
Security Audit — agent-trust-hub — happyhorse