The Agent Skills Directory

[SAFE]: The skill provides instructions and usage examples for the belt CLI tool, allowing the agent to perform video generation tasks through an external service. No malicious patterns were detected in the instructions or command templates.
[EXTERNAL_DOWNLOADS]: The skill references installation instructions hosted on GitHub (inference-sh/skills) and suggests the installation of a related utility skill (belt-sh/cli). These references target the official resources for the platform's ecosystem and are documented neutrally.
[INDIRECT_PROMPT_INJECTION]: The skill provides templates that ingest user-supplied prompts and URLs which are then passed to an external CLI tool. While this represents a potential injection surface for the downstream AI models, the skill itself does not facilitate exploitation, and standard guardrails on the inference platform are expected to handle content safety.
Ingestion points: User-provided prompts, image URLs, and audio URLs within the belt app run command examples in SKILL.md.
Boundary markers: Absent in the provided command templates.
Capability inventory: The skill uses Bash(belt *) to execute commands.
Sanitization: None performed within the skill instructions.

ai-video-generation