The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions for the agent to download and install the belt CLI tool and individual AI applications from the inference-sh organization on GitHub, which is the official source for this service.
[REMOTE_CODE_EXECUTION]: Uses npx and the belt CLI's app run command to fetch and execute remote code logic for interacting with language models.
[COMMAND_EXECUTION]: Relies on the Bash tool to execute CLI commands for authentication, model selection, and prompt execution.
[INDIRECT_PROMPT_INJECTION]: The core functionality involves ingesting user-provided prompts and sending them to external models. This represents an inherent indirect prompt injection surface where unvalidated data is processed by the AI system.

llm-models