TmuxAI is an AI-powered terminal assistant designed for developers and power users who want intelligent pair programming directly within their tmux sessions. It non-intrusively observes terminal content across panes to provide context-aware assistance, execute commands, and suggest workflow improvements without disrupting the user's current setup.

How It Works

TmuxAI operates by observing the visible content of all tmux panes within a window, including current commands, shell type, and OS. It communicates via a dedicated chat pane and executes commands in a separate pane, mimicking human pair programming. This approach allows it to understand the user's workflow and provide assistance by processing user messages along with pane context and chat history sent to an AI model.

Quick Start & Requirements

• Install: curl -fsSL https://get.tmuxai.dev | bash or brew install tmuxai
• Prerequisites: tmux
• Configuration: Set TMUXAI_OPENROUTER_API_KEY environment variable.
• Docs: get.tmuxai.dev

Highlighted Details

• Modes: Supports "Observe Mode" (default), "Prepare Mode" (enhanced command tracking), and "Watch Mode" (proactive suggestions).
• Context Management: Features "squashing" to summarize chat history and manage token usage for AI models.
• AI Provider Flexibility: Configurable to use OpenRouter, OpenAI, Anthropic, or local Ollama endpoints.
• Customization: Offers session-specific configuration overrides via /config command.

Maintenance & Community

• Contributing: Open to pull requests and issues. Project has a GitHub repository.
• Links: No explicit community links (Discord/Slack) or roadmap mentioned in the README.

Licensing & Compatibility

• License: Apache License 2.0.
• Compatibility: Permissive license suitable for commercial use and integration with closed-source projects.

Limitations & Caveats

Prompts are currently tuned for Gemini 2.5, and behavior with other models may vary. The README does not detail specific performance benchmarks or extensive testing across all supported AI models.