Skip to main content
The easiest way to start building agents and applications powered by LLMs—with built-in capabilities for task planning, file systems for context management, subagent-spawning, and long-term memory. You can use deep agents for any task, including complex, multi-step tasks. We think of deepagents as an “agent harness”. It is the same core tool calling loop as other agent frameworks, but with built-in tools and capabilities. deepagents is a standalone library built on top of LangChain’s core building blocks for agents and using LangGraph’s tooling for running agents in production. The deepagents library contains:
  • Deep Agents SDK: A package for building agents that can handle any task
  • Deep Agents CLI: A terminal coding agent built on the Deep Agents SDK
  • ACP integration: An Agent Client Protocol connector for using deep agents in code editors like Zed
LangChain is the framework that provides the core building blocks for your agents. To learn more about the differences between LangChain, LangGraph, and Deep Agents, see Frameworks, runtimes, and harnesses.

​
Create a deep agent

import * as z from "zod";
// npm install deepagents langchain @langchain/core
import { createDeepAgent } from "deepagents";
import { tool } from "langchain";

const getWeather = tool(
  ({ city }) => `It's always sunny in ${city}!`,
  {
    name: "get_weather",
    description: "Get the weather for a given city",
    schema: z.object({
      city: z.string(),
    }),
  },
);

const agent = createDeepAgent({
  tools: [getWeather],
  system: "You are a helpful assistant",
});

console.log(
  await agent.invoke({
    messages: [{ role: "user", content: "What's the weather in Tokyo?" }],
  })
);
See the Quickstart and Customization guide to get started building your own agents and applications with Deep Agents.
Use LangSmith to trace requests, debug agent behavior, and evaluate outputs. Set LANGSMITH_TRACING=true and your API key to get started.

​
When to use the Deep Agents

Use the Deep Agents SDK when you want to build agents that can:
  • Handle complex, multi-step tasks that require planning and decomposition
  • Manage large amounts of context through file system tools and auto-summarization
  • Swap filesystem backends to use in-memory state, local disk, durable stores, sandboxes, or your own custom backend
  • Execute shell commands via the execute tool when using a sandbox backend
  • Delegate work to specialized subagents for context isolation
  • Persist memory across conversations and threads
  • Require human approval for sensitive operations with human-in-the-loop workflows
  • Use any model that supports tool calling — provider agnostic across frontier and open models
For building simpler agents, consider using LangChain’s createAgent or building a custom LangGraph workflow. Use the Deep Agents CLI when you want a coding agent on the command line, built on the Deep Agents SDK:
  • Run interactively or non-interactively — use the CLI as a chat-style coding agent, or pipe tasks with -n for scriptable, headless execution.
  • Customize agents with skills and memory.
  • Teach agents as you use them about your preferences, common patterns, and custom project knowledge.
  • Execute code on your machine or in sandboxes.
  • Extend with MCP tools from external Model Context Protocol servers.

​
Core capabilities

Planning and task decomposition

Deep Agents include a built-in write_todos tool that enables agents to break down complex tasks into discrete steps, track progress, and adapt plans as new information emerges.

Context management

File system tools (ls, read_file, write_file, edit_file) allow agents to offload large context to in-memory or filesystem storage, preventing context window overflow and enabling work with variable-length tool results. Auto-summarization compacts older conversation messages when the context window grows long, keeping the agent effective across extended sessions.

Shell execution

When using a sandbox backend, agents get an execute tool to run shell commands for tests, builds, git operations, and system tasks. Sandbox backends provide isolation so agents can execute code without compromising your host system.

Pluggable filesystem backends

The virtual filesystem is powered by pluggable backends that you can swap to fit your use case. Choose from in-memory state, local disk, LangGraph store for cross-thread persistence, sandboxes for isolated code execution (Modal, Daytona, Deno), or combine multiple backends with composite routing. You can also implement your own custom backend.

Subagent spawning

A built-in task tool enables agents to spawn specialized subagents for context isolation. This keeps the main agent’s context clean while still going deep on specific subtasks.

Long-term memory

Extend agents with persistent memory across threads using LangGraph’s Memory Store. Agents can save and retrieve information from previous conversations.

Human-in-the-loop

Configure human approval for sensitive tool operations using LangGraph’s interrupt capabilities. Control which tools require confirmation before execution.

Skills

Extend agents with reusable skills that provide specialized workflows, domain knowledge, and custom instructions.

Smart defaults

Ships with opinionated system prompts that teach the model how to use its tools effectively — plan before acting, verify work, and manage context. Customize or replace the defaults as needed.

​
Get started

SDK Quickstart

Build your first deep agent

Customization

Learn about customization options for the SDK

Models

Configure models and providers

Backends

Choose and configure pluggable filesystem backends

Human-in-the-loop

Configure approval for sensitive operations

CLI

Use the Deep Agents CLI

Reference

See the deepagents API reference
Edit this page on GitHub or file an issue.
Connect these docs to Claude, VSCode, and more via MCP for real-time answers.