Core Features

• Terminal Control: Convert goals into concrete shell commands with readable explanations and expected outcomes.
• Safe Approval Mode: Approve each command before it runs. Stay in control of what executes and when.
• Auto-Accept Mode: When you trust the plan, let it run. Switch back to review mode anytime.
• Local LLM Support: Use local models through OpenAI-compatible endpoints (LM Studio or your own server).
• Key Management: API keys are stored locally (macOS Keychain) — never in browser storage.
• Troubleshooting Assistant: Paste errors, inspect output, and get step-by-step fixes tailored to your environment.
• Secure Local Execution: Commands run on your machine. You stay in control.

Supported Providers

• OpenAI: Use your OpenAI API key and model selection.
• Anthropic: Bring your Claude key for high-quality reasoning.
• LM Studio: Run models locally via an OpenAI-compatible server.
• Local servers: Point to your own endpoint on 127.0.0.1.

Why not just ChatGPT?

• No copy-pasting commands — AI Terminal generates executable plans
• Local app + your choice of provider (cloud or local)
• Works inside your terminal workflow with direct system context
• Approve before anything runs (safe mode)