The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references installation instructions hosted on GitHub (raw.githubusercontent.com/inference-sh/skills) and suggests installing the belt-sh/cli and related skills via npx. These resources are linked to the official ecosystem of the tool described.
[INDIRECT_PROMPT_INJECTION]: The skill demonstrates constructing shell commands (belt app run) using JSON payloads that include user-defined prompts. If these commands are executed with untrusted data without proper sanitization, they could be susceptible to indirect prompt injection within the target AI model's context.
Ingestion points: Prompt field within the JSON input for the belt command.
Boundary markers: None explicitly defined in the examples.
Capability inventory: Network communication with the inference.sh API via the belt tool.
Sanitization: Not present in the demonstrated command templates.

video-prompting-guide