The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: References installation guides and related components on the inference-sh GitHub repository for the purpose of setting up the CLI.
[COMMAND_EXECUTION]: Documents the use of the belt CLI utility for authentication and running model inferences, which is consistent with the skill's stated purpose.
[PROMPT_INJECTION]: Documentation includes examples of passing user-defined prompts to LLMs via the --input flag, representing an indirect prompt injection surface.
Ingestion points: SKILL.md examples using the belt app run command.
Boundary markers: User content is structured within JSON prompt fields in provided examples.
Capability inventory: Shell execution of the belt CLI utility for remote model interaction.
Sanitization: None provided within the documentation skill itself.

llm-models