Description
Openvibe is a Python implementation of an open-source AI coding agent, offering a modular framework for Auto-GPT development. It allows users to build and experiment with AI agents that can perform coding tasks. The framework supports various LLM backends through its LLM abstraction layer, providing flexibility in model selection. It also includes built-in tools for common tasks like bash execution, file reading and writing, and web fetching.
Read More
The architecture of openvibe is centered around key modules, including SQLite for database abstraction, LLM abstraction via litellm, and a FastAPI-based server for HTTP and SSE communication. The session/processor module handles the core agent execution loop, while the tool directory houses the built-in tools. Configuration is managed through a `openvibe.json` file, allowing users to specify model providers, model IDs, and agent permissions. The framework supports session management, enabling users to list and inspect active sessions.
Openvibe's value proposition lies in its modularity and extensibility. Developers can easily swap out components like the database or LLM backend. The framework's built-in tools and session management features streamline the development and deployment of AI coding agents. The project is hosted on GitHub, with active contributions from a community of developers. The project's documentation provides clear instructions for installation, usage, and configuration, making it accessible to both beginners and experienced developers. The framework's focus on open-source principles and community contributions fosters collaboration and innovation in the field of AI coding agents.
The target audience includes developers, researchers, and AI enthusiasts interested in building and experimenting with AI coding agents. The framework's modular design and clear documentation make it suitable for both prototyping and production use. The availability of built-in tools and session management features simplifies the development process, allowing users to focus on the core logic of their AI agents. The project's open-source nature encourages community contributions and collaboration, fostering a dynamic and evolving ecosystem.
- CategoryAI Agent Builders
- Pricing ModelOpen source
- DeploymentCloud, Self-hosted
- PlatformsWeb
openvibe's Core Features
Modular architecture for Auto-GPT development
LLM abstraction layer with litellm backend
Built-in tools for bash, read, write, edit, and more
FastAPI HTTP + SSE server for communication
Session management for agent control
SQLite database abstraction
Configuration via openvibe.json
Open-source and community-driven
Supports Anthropic and other LLM providers
Easy installation via pip
Integration with MCP server
Core agent execution loop
Getting Started with openvibe
Clone: Clone the openvibe repository from GitHub.
Install: Install the framework and dependencies using pip: `pip install -e .[dev]`
Configure: Create an `openvibe.json` file in your project root to configure the model and permissions.
Set API Key: Set your API key as an environment variable: `export ANTHROPIC_API_KEY=sk-...`
Execute: Run the server using `openvibe serve` or execute a one-shot prompt with `openvibe run "fix the failing tests"`.
Manage Sessions: Use commands like `openvibe session list` and `openvibe session show <session-id>` to manage sessions.
Explore: Review the source code in `src/openvibe/` to understand the architecture.
openvibe's Use Cases
- Automated Coding
- Test Automation
- Code Debugging
- Software Prototyping
- Code Refactoring
- AI Agent Development
