[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"readme:deepagents":3},"\u003Cdiv align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fdocs.langchain.com\u002Foss\u002Fpython\u002Fdeepagents\u002Foverview#deep-agents-overview\" rel=\"nofollow ugc noopener\">\n    \u003Cpicture>\n      \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\".github\u002Fimages\u002Flogo-dark.svg\">\u003C\u002Fsource>\n      \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\".github\u002Fimages\u002Flogo-light.svg\">\u003C\u002Fsource>\n      \u003Cimg alt=\"Deep Agents Logo\" src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flangchain-ai\u002Fdeepagents\u002FHEAD\u002F.github\u002Fimages\u002Flogo-dark.svg\" width=\"50%\" \u002F>\n    \u003C\u002Fpicture>\n  \u003C\u002Fa>\n\u003C\u002Fdiv>\u003Cdiv align=\"center\">\n  \u003Ch3>The batteries-included agent harness.\u003C\u002Fh3>\n\u003C\u002Fdiv>\u003Cdiv align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fopensource.org\u002Flicenses\u002FMIT\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fl\u002Fdeepagents\" alt=\"PyPI - License\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fpypistats.org\u002Fpackages\u002Fdeepagents\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fpepy\u002Fdt\u002Fdeepagents\" alt=\"PyPI - Downloads\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fpypi.org\u002Fproject\u002Fdeepagents\u002F#history\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fv\u002Fdeepagents?label=%20\" alt=\"Version\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fx.com\u002Flangchain_oss\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Ftwitter\u002Furl\u002Fhttps\u002Ftwitter.com\u002Flangchain_oss.svg?style=social&label=Follow%20%40LangChain\" alt=\"Twitter \u002F X\" \u002F>\u003C\u002Fa>\n\u003C\u002Fdiv>\u003Cbr \u002F>\u003Cp>Deep Agents is an open source agent harness — an opinionated agent that runs out of the box. Extend, override, or replace any piece.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Principles:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Opinionated\u003C\u002Fstrong> — defaults tuned for long-horizon, multi-step work\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Extensible\u003C\u002Fstrong> — override or replace any piece without forking\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Model-agnostic\u003C\u002Fstrong> — works with any LLM that supports tool calling: frontier, open-weight, or local\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Production-ready\u003C\u002Fstrong> — built on LangGraph (streaming, persistence, checkpointing) with first-class tracing, evaluation, and deployment via LangSmith\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cstrong>Features include:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Sub-agents\u003C\u002Fstrong> — delegate tasks to agents with isolated context windows\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Filesystem\u003C\u002Fstrong> — read, write, edit, or search over pluggable local, sandboxed, or remote backends\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Context management\u003C\u002Fstrong> — summarize long threads and offload tool outputs to disk\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Shell access\u003C\u002Fstrong> — run commands in your sandbox of choice\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Persistent memory\u003C\u002Fstrong> — pluggable state and store backends for cross-session recall\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Human-in-the-loop\u003C\u002Fstrong> — approve, edit, or reject tool calls before they run\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Skills\u003C\u002Fstrong> — reusable behaviors the agent can load on demand\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Tools\u003C\u002Fstrong> — bring your own functions or any MCP server\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Deep Agents is available as a JavaScript\u002FTypeScript library — see \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fdeepagentsjs\" rel=\"nofollow ugc noopener\">deepagents.js\u003C\u002Fa>.\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>[!NOTE]\n\u003Cstrong>Deep Agents Code\u003C\u002Fstrong> — a pre-built coding agent in your terminal, similar to Claude Code or Cursor, powered by any LLM. Install with \u003Ccode>curl -LsSf https:\u002F\u002Flangch.in\u002Fdcode | bash\u003C\u002Fcode>. See the \u003Ca href=\"https:\u002F\u002Fdocs.langchain.com\u002Fdeepagents-code\" rel=\"nofollow ugc noopener\">documentation\u003C\u002Fa> for the full feature set.\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Ch2>Quickstart\u003C\u002Fh2>\n\u003Cpre>\u003Ccode class=\"language-bash\">uv add deepagents\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cpre>\u003Ccode class=\"language-python\">from deepagents import create_deep_agent\n\nagent = create_deep_agent(\n    model=\"openai:gpt-5.5\",\n    tools=[my_custom_tool],\n    system_prompt=\"You are a research assistant.\",\n)\nresult = agent.invoke({\"messages\": \"Research LangGraph and write a summary\"})\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>The agent can plan, read\u002Fwrite files, and manage its own context. Add your own tools, swap models, customize prompts, configure sub-agents, and more. See the \u003Ca href=\"https:\u002F\u002Fdocs.langchain.com\u002Foss\u002Fpython\u002Fdeepagents\u002Foverview\" rel=\"nofollow ugc noopener\">documentation\u003C\u002Fa> for full details.\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>[!TIP]\nFor developing, debugging, and deploying AI agents and LLM applications, see \u003Ca href=\"https:\u002F\u002Fdocs.langchain.com\u002Flangsmith\u002Fhome\" rel=\"nofollow ugc noopener\">LangSmith\u003C\u002Fa>.\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Ch2>FAQ\u003C\u002Fh2>\n\u003Ch3>How is this different from LangGraph or LangChain?\u003C\u002Fh3>\n\u003Cp>LangGraph is the graph runtime. LangChain's \u003Ccode>create_agent\u003C\u002Fcode> is a minimal agent harness on top of it. Deep Agents is a more opinionated harness on top of \u003Ccode>create_agent\u003C\u002Fcode> — same building blocks, but with filesystem, sub-agents, context management, and skills bundled in. For how the three relate, see the \u003Ca href=\"https:\u002F\u002Fdocs.langchain.com\u002Foss\u002Fpython\u002Fconcepts\u002Fproducts\" rel=\"nofollow ugc noopener\">LangChain ecosystem overview\u003C\u002Fa>.\u003C\u002Fp>\n\u003Ch3>Does this work with open-weight or local models?\u003C\u002Fh3>\n\u003Cp>Yes. Any model \u003C\u002Fp>\n",1784240407114]