OpenMuse is a personal agent framework that integrates a browser, terminal, and file management capabilities, designed to facilitate ongoing tasks and workflows. Built with CopilotKit and AG-UI, it supports various interactions, including task delegation and real-time browsing.
OpenBot is a local-first desktop application designed to run AI agents collaboratively on your computer. It supports various AI models, including Codex, Claude, and Grok, providing dedicated workspaces, task queues, file sharing, and agent-to-agent communication.
treg.to is an OpenRouter platform designed for agent tools that provides access to a curated catalog of over 3,700 endpoints across 99 providers. It enables users to perform various tasks such as SEO, data enrichment, social media analysis, and scraping without needing individual vendor accounts, allowing for a streamlined approach to accessing diverse data services.
Buzz is a hive mind communication platform designed for collaborative work between humans and AI agents. It enables users to manage workflows, triage bugs, and maintain an audit trail, all while providing a unified space for discussions, code patches, and approvals.
Executor is an open-source integration layer designed for AI agents, enabling them to connect with various APIs such as OpenAPI, MCP servers, and GraphQL in a secure environment. It allows for the configuration of integrations with authentication and per-tool policies, facilitating a unified catalog for any MCP-compatible agent.
Browser Harness is a self-healing framework that connects large language models (LLMs) directly to a real browser, enabling them to perform tasks with complete freedom. It utilizes a thin, editable Chrome DevTools Protocol (CDP) harness that allows the agent to adapt and improve its execution capabilities over time.
rag.computer is an open-source, self-hostable RAG platform that facilitates document ingestion, embedding, and vector search through a straightforward REST API. It supports various document formats and offers capabilities for semantic, keyword, and hybrid search using advanced embedding providers.
Centaur is a production control plane designed for shared AI agents that operate in isolated sandboxes and utilize approved tools. It enables teams to interact with agents via Slack, allowing them to execute real tasks in a secure environment with access to various development tools.
OpenClicky is an open-source macOS menu-bar companion that hosts Clicky, an AI assistant providing voice help, screen-aware responses, and visual results through an image gallery. It operates with local configuration, eliminating the need for Google login or hosted key-sync flows.
browse.sh is a scriptable browsing workspace designed for fast web access and automation. It serves as a collection of agent skills that enable Claude Code to interact with the web through browser automation using the official browse CLI.
AG-UI is an open, lightweight, event-based protocol designed to standardize the connection between AI agents and user-facing applications. It facilitates seamless integration of AI agents with real-time user context and user interfaces, enabling features like bi-directional state synchronization and generative UI.
watch-cli is a tool that enables AI agents to process and analyze social videos by extracting video files, evenly-spaced frames, and full audio transcripts. It supports various platforms including YouTube, LinkedIn, TikTok, and more, and is designed to be cost-effective compared to traditional multimodal APIs.
Gentle-AI is an ecosystem configurator designed to enhance AI coding agents by providing features such as persistent memory, Spec-Driven Development workflows, curated coding skills, and an AI provider switcher. It allows for per-phase model assignment, enabling different models to be used for each step of the development process.
svelte-edge is a demonstration project that allows a chat agent to generate Svelte 5 user interfaces, which are compiled at the edge using Cloudflare Workers. The generated components can be integrated inline within conversations and can communicate structured data back to the host page, enabling seamless workflow continuation.
The most popular framework for developing LLM-powered applications. Includes agents, chains, memory, and integrations.
Self-hosted AI interface compatible with Ollama and OpenAI APIs. Feature-rich ChatGPT-like UI for local models.
Open-source platform to build and operate LLM apps. Visual workflow builder, RAG pipeline, and agent orchestration.
Open-source browser automation for AI agents. Works with LangChain and other frameworks out of the box.
Search API optimized for AI agents. Returns clean, structured results tailored for LLM consumption.
Durable execution for long-running agent workflows. Handles retries, timeouts, and state persistence automatically.
Workflow orchestration for AI pipelines. Build, schedule, and monitor agentic workflows with observability built in.
Platform for software development agents. Agents write code, run commands, and browse the web autonomously.
Data framework for LLM applications. Ingests, structures, and accesses private or domain-specific data for agents.
Microsoft's multi-agent conversation framework. Agents customize, converse with each other, and autonomously execute tasks.
Framework for orchestrating role-playing, autonomous AI agents. Enables collaborative multi-agent workflows with ease.