Harnesses & SDKs

Open-source Python framework for multi-agent workflows, with role-based Crews and event-driven Flows.

SDK for browser agents with act, observe and extract calls, used from TypeScript and Python to automate sites and pull out structured data.

Python SDK that runs the Claude Code agent loop inside your own program, with custom tools and hooks written as Python functions.

Open-source framework that routes each user query to one of several AI agents, with Python, TypeScript and Swift runtimes.

Python framework and runtime for building agents, serving them over a REST API and MCP server, and managing them from a web control plane.

TypeScript toolkit for AI apps and agents with a unified provider API, streaming, UI hooks and a wrapper for running established agent harnesses.

Framework for Python or Node.js programs that join LiveKit rooms as realtime voice and video agents.

Open-source agent harness from LangChain for building agents with code, with LangSmith for testing and monitoring.

Open-source Python framework for real-time voice and multimodal agents, built from composable pipelines with handoffs between specialist agents.

JavaScript and TypeScript framework of interoperable components for building LLM apps and agents, with LangGraph.js and Deep Agents above it.

TypeScript SDK that gives programs Claude Code's built-in tools for reading, editing and running code, installed from npm.

Open-source Python framework from deepset for modular pipelines and agent workflows covering retrieval, tool use and generation.

Small Python library whose agents write their actions as code, with sandboxed execution and tools from MCP servers, LangChain or the Hub.

TypeScript platform to build and deploy long-running AI agents and workflows with retries, queues and human approval steps.

Browser DOCX editor and document engine built on OOXML, with a Document API that agents reach through SDKs, a CLI and an MCP server.

Coinbase toolkit that gives AI agents a crypto wallet and onchain actions, framework- and wallet-agnostic.

Go framework for LLM applications with chat-model agents, multi-agent coordination, graph workflows and human-in-the-loop resume.

Java library with one API over LLM providers and vector stores, plus tools, RAG and integrations for Quarkus and Spring Boot.

Java framework for agentic, workflow and multi-agent AI apps, with graph workflows, MCP and A2A support.

Python agent framework from Alibaba that lets the model drive tool use, with team pipelines, MCP, agent skills, A2A, memory and tracing.

TypeScript framework for AI agents with tools, memory, RAG, MCP and sub-agents, paired with a Cloud or self-hosted console.

Python library that builds a tree index of long documents and has an LLM reason through it to retrieve answers, without a vector database or chunking.

AI application runtime framework for backends with structured outputs, observable actions, MCP capabilities and recoverable workflows.

Rails framework for building AI agents as controllers, with provider gems for OpenAI, Anthropic, Ollama and more.