Can I use Command completely offline?

Yes! Command supports Ollama for fully local execution with zero external API calls. Run models like Llama 3.1, Mistral NeMo, Gemma 2, or Qwen 2 entirely on your laptop with complete privacy.

How does Command integrate with Obsidian and VS Code?

Command works seamlessly with markdown content. You can create custom workflows in command.yml that reference markdown files in your Obsidian vault or VS Code workspace. Use the 'refer' command to process markdown content with AI models.

Terminal Productivity With 15+ AI Models

Multi-provider CLI with markdown automation, Obsidian integration, and 15+ models. Fork this to build CLI tools, DevOps automation, and developer productivity apps.

View Code

You're Building CLI Tools, But...

Multi-provider support is tedious. Markdown workflows take time. Managing 15+ models requires boilerplate code.

Building from Scratch

3-6 months

✗ Wire up 7 AI providers
✗ Build CLI interface from scratch
✗ Create workflow automation
✗ Debug edge cases

Single-Provider CLIs

Limited

✗ Locked to one provider
✗ No workflow automation
✗ Basic chat only
✗ Can't customize easily

REFERENCE READY

Fork Command

Ship in 2-7 Days

✓ 7 providers, 15+ models ready
✓ Markdown workflows built-in
✓ Audit trails + config system
✓ Fork, customize, ship

Fork This To Build

DevOps Automation Tools

CI/CD assistants, deployment tools, infrastructure automation

CLI Applications

Developer tools, system utilities, command-line interfaces

Workflow Managers

Task automation, pipeline orchestration, workflow engines

Terminal AI Assistants

Coding assistants, debug helpers, terminal copilots

Code Generation CLIs

Boilerplate generators, scaffolding tools, code assistants

Documentation Generators

API docs, code comments, readme automation, wiki generators

Task Automation Systems

Batch processing, file manipulation, system administration

Developer Productivity Tools

Git helpers, terminal utilities, workflow accelerators

Working Features

Latest Models Built-in

Access the newest Claude 4.5, GPT-4o, and Gemini 1.5 models with simple aliases.

• Claude Sonnet 4.5, Haiku 4.5, Opus 4.1
• OpenAI GPT-4o, GPT-4o Mini
• Google Gemini 1.5 Pro/Flash

Effortless Setup

Get started in 60 seconds with simple .env.local configuration.

pip install command
cmnd init
ask "Hello!"

Privacy Controls

Choose fully local execution or trusted cloud providers.

• Fully local with Ollama
• Hybrid workflows
• Your data, your choice

Core Features

Markdown Workflows

Create custom workflows that process markdown content. Perfect for Obsidian and VS Code users.

Model Testing

Compare response quality, latency, and cost across all 15+ models with a single command.

Performance Trending

Visualize latency and token trends over time with built-in sparkline charts.

Intent-Driven

Manage tasks, goals, and plans as markdown outlines. Your intents power your workflow.

Audit Trail

Every command, prompt, and response logged in trail.yml for complete transparency.

Configurable

Customize everything via command.yml. Define your own workflows, models, and shortcuts.

Installation

1. Install Command

pip install command

2. Initialize

cmnd init

Creates command.yml and .env.local.example in current directory

3. Configure API Keys

Copy .env.local.example to .env.local and add your keys:

ANTHROPIC_API_KEY=your_key
OPENAI_API_KEY=your_key
GOOGLE_API_KEY=your_key

4. Start Using

cmnd id # Show active provider and model
ask "How old is the oldest pyramid?"

FAQ

What is Command?

Command is an AI-powered terminal productivity tool that gives you access to 15+ leading AI models from 7 providers directly in your command line.

Which models are supported?

Claude (Sonnet 4.5, Haiku 4.5, Opus 4.1), OpenAI (GPT-4o, GPT-4o Mini), Google (Gemini 1.5 Pro/Flash), and local models via Ollama (Llama 3.1, Mistral NeMo, Gemma 2, Qwen 2).

Can I use Command offline?

Yes! Command supports Ollama for fully local execution with zero external API calls. Complete privacy on your laptop.

How does it integrate with Obsidian?

Command processes markdown files seamlessly. Create custom workflows in command.yml that reference your Obsidian vault. Use the 'refer' command to process notes with AI.

Learn More About Command

Supercharge Your Terminal with AI

Turn your Terminal into a configurable, interactive AI workspace with 15+ LLMs and 7 providers at your fingertips. Command is all you need.

Read the full article

Markdown Workflows for Content Creation

Learn how Command transforms partial ideas into polished content through intent-driven workflows that integrate seamlessly with your favorite markdown tools.

Read the full article

Build CLI Tools That Ship Fast

Fork Command and ship DevOps automation, terminal assistants, or productivity CLIs in days. 7 providers, 15+ models, battle-tested.

View Code