OPEN-SOURCE AI AGENT DIRECTORY

Browser interface for autonomous task agents; repository archived and setup may require older dependencies.

Local-first autonomous assistant for browsing, coding, planning, research, and voice; hardware intensive.

Python multi-agent framework with tools and console examples for building and running agent workflows.

AutoAgent, formerly known as MetaChain, provides a natural-language interface for creating and customizing AI agents, tools, and workflows with Docker-backed execution.

Python browser automation agent; local OSS package plus optional paid cloud.

Open-source environment and benchmark suite for testing browser-control agents across web tasks.

Terminal coding assistant that maps repositories, edits code, runs commands and integrates with Git.

Multi-agent software development system that coordinates roles to design and build applications.

Terminal-based coding agent for repository tasks and Git workflows.

User-facing coding agent with file operations, terminal and browser tools, and iterative approval loops.

Coding agent with a core binary, optional execution tools and a local web interface.

OpenAI terminal coding agent for software engineering tasks.

Document/RAG workspace with agents and a visual no-code Agent Flow editor.

Desktop AI workspace with assistants, knowledge bases, MCP and agents.

Node-based visual graph engine and API for modular image/video generation workflows.

Document-grounded agent workspace with research features, drag-and-drop workflow builder, APIs and integrations.

Document-grounded chat and audio-summary workspace template assembled with Supabase, Bolt, n8n and Netlify.

Image generation and editing/canvas application with optional node workflows.

Research framework for multi-agent simulations, collaboration, and evaluation with included example scenarios.

Data-assistant workspace and framework for building data agents/apps, with web UI and database/knowledge-base tools.

Node.js research CLI that searches the web, processes results and writes Markdown reports.

Web research application that searches sources and produces cited reports with evidence inspection and follow-up.

Self-hosted finance-specific research agent with terminal UI and optional WhatsApp gateway.

Programming framework for LM pipelines, retrieval systems, and agent-loop optimization.

Browser interface for autonomous task agents; repository archived and setup may require older dependencies.

Local-first autonomous assistant for browsing, coding, planning, research, and voice; hardware intensive.

Python multi-agent framework with tools and console examples for building and running agent workflows.

AutoAgent, formerly known as MetaChain, provides a natural-language interface for creating and customizing AI agents, tools, and workflows with Docker-backed execution.

Python browser automation agent; local OSS package plus optional paid cloud.

Open-source environment and benchmark suite for testing browser-control agents across web tasks.

Terminal coding assistant that maps repositories, edits code, runs commands and integrates with Git.

Multi-agent software development system that coordinates roles to design and build applications.

Terminal-based coding agent for repository tasks and Git workflows.

User-facing coding agent with file operations, terminal and browser tools, and iterative approval loops.

Coding agent with a core binary, optional execution tools and a local web interface.

OpenAI terminal coding agent for software engineering tasks.

Document/RAG workspace with agents and a visual no-code Agent Flow editor.

Desktop AI workspace with assistants, knowledge bases, MCP and agents.

Node-based visual graph engine and API for modular image/video generation workflows.

Document-grounded agent workspace with research features, drag-and-drop workflow builder, APIs and integrations.

Document-grounded chat and audio-summary workspace template assembled with Supabase, Bolt, n8n and Netlify.

Image generation and editing/canvas application with optional node workflows.

Research framework for multi-agent simulations, collaboration, and evaluation with included example scenarios.

Data-assistant workspace and framework for building data agents/apps, with web UI and database/knowledge-base tools.

Node.js research CLI that searches the web, processes results and writes Markdown reports.

Web research application that searches sources and produces cited reports with evidence inspection and follow-up.

Self-hosted finance-specific research agent with terminal UI and optional WhatsApp gateway.

Programming framework for LM pipelines, retrieval systems, and agent-loop optimization.