64 harnesses
Run large language models locally with a one-line install and a simple CLI.
Microsoft's Python utility for converting documents (PDF, DOCX, PPTX, etc.) into Markdown for LLMs.
Make websites accessible to AI agents — the easiest way to give LLMs browser superpowers.
The foundational framework for building applications powered by LLMs with chains, agents, and tools.
Cross-browser end-to-end testing and automation framework from Microsoft.
Open-source LLM-friendly web crawler and scraper optimized for RAG.
VS Code-integrated autonomous coding agent that controls files, terminal, and browser.
Multi-agent conversation framework from Microsoft for solving complex tasks with collaborating AIs.
Generate an entire codebase from a natural-language project description.
Framework for orchestrating role-based AI agent crews that collaborate on tasks.
Local-first autonomous AI agent with MCP (Model Context Protocol) integration.
Terminal-based AI pair programmer with native Git integration and automatic commits.
Lightweight, opinionated framework for building production-ready AI agents and assistants.
Data framework for building RAG and agentic applications over private data.
Browser-based platform for assembling and deploying autonomous AI agents.
High-throughput, memory-efficient inference and serving engine for LLMs.
Open-source AI code assistant for VS Code and JetBrains with custom model and tool support.
Stateful, graph-based framework for orchestrating complex multi-agent LLM workflows.
Microsoft's SDK for integrating LLMs into apps with planners, skills, and memory.
Tiny, opinionated agent framework from Hugging Face — agents that write code as actions.
OpenAI's official Python SDK for building production multi-agent workflows.
Open-source framework for building production-grade LLM applications and RAG pipelines.
Memory layer for AI agents — adds long-term, personalized memory to LLM applications.
LLM-as-controller system that orchestrates Hugging Face models for multimodal tasks.