jervis-llm

Universal local LLM lifecycle CLI.

Quick start

1. Install dependencies:

npm install

2. Build the CLI:

npm run build

3. Run environment checks:

jervis-llm doctor
jervis-llm --doctor
jlm --doctor

4. Get recommendations and run-fit:

jervis-llm recommend --profile coding
jervis-llm can-run qwen2.5-coder:7b --provider ollama

5. Install, run, inspect running models, and stop:

jervis-llm install qwen2.5-coder:7b --provider ollama
jervis-llm run qwen2.5-coder:7b --provider ollama
jervis-llm running
jervis-llm stop qwen2.5-coder:7b --provider ollama
jervis-llm stop --all --provider ollama

Global install and usage

You can install the CLI globally and use it from anywhere:

npm install -g jervis-llm
# Full name
jervis-llm --doctor
jervis-llm recommend --profile coding
# Or with the short alias
jlm --doctor
jlm recommend --profile coding

If you are developing locally, use npm run dev -- ... as above.

Examples:

npm run dev -- doctor
npm run dev -- running
npm run dev -- stop --all --provider ollama

Documentation site

• Start docs locally:
  • npm run docs:dev
• Build docs:
  • npm run docs:build
• Preview built docs:
  • npm run docs:preview

Docs source lives in docs/.

Common command patterns

# List installed models
jervis-llm models

# List currently running models
jervis-llm running

# Check if a specific model can run on this machine
jervis-llm can-run qwen3.6:27b --provider ollama

# Stop one running model or all running models
jervis-llm stop qwen3.6:27b --provider ollama
jervis-llm stop --all --provider ollama

Documentation

Full documentation available at the Jervis LLM Docs:

License

MIT © Kerolos Zakaria