llm - Drive language models via Simon Willison's llm CLI
Executes LLM workflows via shell CLI including prompts, chat, file attachments, structured schemas, fragments, plugins, and log management.
Tags
Updated: 2026-09-18Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Execute prompt and chat workflows
- Attach local files and URLs
- Generate structured output using schemas
- Manage context fragments and aliases
- Install and manage CLI plugins
- Query and inspect SQLite logs
Inputs
- Prompts and system instructions
- Local files, STDIN, and URLs
- JSON schema specifications
- LLM provider API keys
Outputs
- Model responses and text streams
- Structured JSON output
- SQLite database log entries
- Saved templates and fragment aliases
Requirements
- llm CLI
- Python package manager uv, pipx, or brew
- LLM provider API key
- Schema or multimodal capable LLM model
