Agent
OpenAI 官方轻量 Agent 框架:以 Agent、Handoff(任务移交)、Guardrail(安全护栏)三个原语构建多 Agent 系统,代码量少、心智负担低,是 Swarm 的生产级继任者。OpenAI's lightweight agent framework built on three primitives — agents, handoffs and guardrails — the production-ready successor to Swarm.
RAG
以数据为中心的 LLM 框架:RAG 索引、检索、文档解析能力业界最全,现已扩展出 Agent 工作流(Workflows),是知识库类应用的首选底座。The data-centric LLM framework: the richest RAG indexing/retrieval toolchain, now extended with agent workflows.
Agent
Pydantic 团队出品的 Agent 框架:用类型注解约束模型输出,把「结构化输出 + 依赖注入 + 可测试」做到极致,工程化程度高,适合严谨的后端项目。Agent framework from the Pydantic team: typed, validated outputs with dependency injection and testability — built for serious backend engineering.
Agent
最流行的 LLM 应用开发框架,提供链式调用、Agent 编排、工具调用与 RAG 全套能力,Python 与 JS 双语言生态,适合快速搭建各类 AI Agent 应用。The most popular framework for building LLM applications: chains, agent orchestration, tool calling and RAG out of the box, in both Python and JavaScript.
编程
让 LLM 在你本机直接写并执行代码的自然语言界面:数据分析、批量文件处理、系统操作,一句话搞定,注意执行前审查代码。A natural-language interface that lets LLMs write and run code locally — data analysis, batch file ops and system tasks.
MCP
最流行的 Python MCP Server 开发框架,几行装饰器即可把函数发布为 MCP 工具,开发效率极高。The popular Pythonic framework for building MCP servers — expose functions as tools with a few decorators.
工程
让 AI Agent 直接「使用浏览器」的 Python 库:视觉+DOM 双通道理解页面,网页操作任务成功率业界领先。Python library that lets agents drive a browser visually and via DOM — state-of-the-art on web task benchmarks.