gptme/gptme
Your agent in your terminal, equipped with local tools: writes code, uses the terminal, browses the web. Make your own persistent autonomous agent on top!
What it solves
gptme is a personal AI agent designed to operate directly within your terminal, allowing it to perform complex tasks like coding, file management, and web browsing. It bridges the gap between a chat-based AI and your actual computer environment, enabling the agent to execute shell commands, run Python scripts, and interact with your desktop as if it were a human user.
How it works
The project acts as a command-line interface (CLI) that connects your preferred AI model (such as Claude, GPT, or local models via Ollama) to your local machine. It provides the AI with a suite of "tools"—such as file editing, web searching, and shell execution—that it can use to complete tasks. It is self-correcting, meaning if a command fails, the agent receives the error output and can attempt to fix the issue. It also supports autonomous operation, where it can run on a schedule to perform background tasks like monitoring GitHub or automating workflows.
Who it’s for
gptme is for developers, power users, and anyone performing knowledge work who wants an AI assistant that can directly manipulate files, run code, and navigate their local environment. It is particularly useful for those who need a persistent agent that can handle long-running tasks on servers, laptops, or CI pipelines.
Highlights
- Broad Compatibility: Works with major AI providers (Anthropic, OpenAI, Google, DeepSeek) and local models via Ollama or OpenAI-compatible servers.
- Extensive Toolset: Includes tools for shell execution, Python, file patching, web browsing, vision, and even full desktop computer use.
- Autonomous Capabilities: Supports persistent, scheduled agents that can manage their own task queues and learn from experience using a "lessons" system.
- Integration Ready: Supports the Model Context Protocol (MCP) and Agent Client Protocol (ACP), allowing it to work as both a client and a server for other development tools like Zed or Claude Desktop.
- Cross-Harness Memory: Maintains a shared memory store that works across different agent tools and harnesses.
Related
- Project
- Project
- Project
- Project
- Project