langchain-ai / deepagents
★ 25,700A skill package that teaches your agent 17 capabilities — every one documented and browsable below, no GitHub required · by langchain-ai.
Each skill below is one capability this package teaches your agent. Install the whole package, or open a skill to install just that one.
Search arXiv for preprints and academic papers by topic with abstract retrieval. Query-based search across physics, mathematics, computer science, biology, statistics, and related fields Configurable result limit (default 10 papers) with results sorted by relevance Returns title and abstract for each matching paper Requires the arxiv Python package; install via pip if not already present
2 files — installable on its own
Long-form blog post writing with research delegation, structured content templates, and AI-generated cover images. Delegates research to subagents before writing, storing findings in markdown for reference and context Enforces a five-part post structure: hook, context, main content (3–5 sections), practical application, and conclusion with call-to-action Generates SEO-optimized cover images using detailed prompts covering subject, style, composition, color, and lighting Outputs posts to...
1 file — installable on its own
Perform a structured code review of changes, checking for correctness, style, tests, and potential issues.
2 files — installable on its own
Read the user's coding preferences from /memory/coding-prefs.md before making non-trivial style decisions, and append new preferences when the user gives…
1 file — installable on its own
When asked to analyze competitors:
1 file — installable on its own
Use for GPU-accelerated data analysis on datasets, CSVs, or tabular data using NVIDIA cuDF. Triggers when tasks involve groupby aggregations, statistical…
1 file — installable on its own
Use for GPU-accelerated machine learning on tabular data using NVIDIA cuML. Triggers when tasks involve classification, regression, clustering, dimensionality…
1 file — installable on its own
Use for creating publication-quality charts and multi-panel analysis summaries. Triggers when tasks involve visualizing data, plotting results, creating…
1 file — installable on its own
Use when processing large PDFs, document collections, or bulk text extraction tasks that benefit from GPU-accelerated processing. Triggers when the user…
1 file — installable on its own
Access LangGraph documentation to build stateful agents and multi-agent workflows. Fetches official LangGraph Python docs covering state machines, graph-based agent design, and human-in-the-loop patterns Prioritizes relevant documentation by query type: implementation guides for how-to questions, concept pages for theory, tutorials for end-to-end examples, and API references for technical details Automatically selects 2–4 most relevant documentation URLs and retrieves their content to answer...
1 file — installable on its own
Break down a coding task into a structured implementation plan with clear steps, file identification, and risk assessment.
1 file — installable on its own
Writes and executes SQL queries from simple SELECTs to complex multi-table JOINs, aggregations, and subqueries. Use when the user asks to query a database,…
1 file — installable on its own
Review the current conversation and capture valuable knowledge — best practices, coding conventions, architecture decisions, workflows, and user feedback —…
1 file — installable on its own
Lists tables, describes columns and data types, identifies foreign key relationships, and maps entity relationships in a database. Use when the user asks about…
1 file — installable on its own
Guide for creating effective skills that extend agent capabilities with specialized knowledge, workflows, or tool integrations. Use this skill when the user…
4 files — installable on its own
Drafts platform-specific social media posts with research-backed content and generated companion images. Supports LinkedIn posts (1,300 characters with professional tone) and Twitter/X threads (280 characters per tweet with 1/🧵 format) Requires delegating research to a subagent before writing, then reading findings to ensure accuracy and relevance Generates eye-catching social images automatically using generate_social_image tool with bold, high-contrast compositions optimized for small...
1 file — installable on its own
Orchestrates multi-source web research by delegating to subagents, synthesizing findings, and producing cited reports. Breaks research questions into 2–5 distinct subtopics, creates a research plan file, and spawns up to 3 subagents in parallel for efficient investigation Each subagent performs 3–5 web searches per subtopic and writes findings to local files with key facts, quotes, and source URLs Synthesizes results by reading local findings files, integrating insights across subtopics, and...
1 file — installable on its own
Deep Agents is an open source agent harness — an opinionated agent that runs out of the box. Extend, override, or replace any piece.
Principles:
- Opinionated — defaults tuned for long-horizon, multi-step work
- Extensible — override or replace any piece without forking
- Model-agnostic — works with any LLM that supports tool calling: frontier, open-weight, or local
- Production-ready — built on LangGraph (streaming, persistence, checkpointing) with first-class tracing, evaluation, and deployment via LangSmith
Features include:
- Sub-agents — delegate tasks to agents with isolated context windows
- Filesystem — read, write, edit, or search over pluggable local, sandboxed, or remote backends
- Context management — summarize long threads and offload tool outputs to disk
- Shell access — run commands in your sandbox of choice
- Persistent memory — pluggable state and store backends for cross-session recall
- Human-in-the-loop — approve, edit, or reject tool calls before they run
- Skills — reusable behaviors the agent can load on demand
- Tools — bring your own functions or any MCP server
Deep Agents is available as a JavaScript/TypeScript library — see deepagents.js.
[!NOTE] Deep Agents Code — a pre-built coding agent in your terminal, similar to Claude Code or Cursor, powered by any LLM. Install with
curl -LsSf https://langch.in/dcode | bash. See the documentation for the full feature set.
Quickstart
uv add deepagentsfrom deepagents import create_deep_agent
agent = create_deep_agent(
model="openai:gpt-5.5",
tools=[my_custom_tool],
system_prompt="You are a research assistant.",
)
result = agent.invoke({"messages": "Research LangGraph and write a summary"})The agent can plan, read/write files, and manage its own context. Add your own tools, swap models, customize prompts, configure sub-agents, and more. See the documentation for full details.
[!TIP] For developing, debugging, and deploying AI agents and LLM applications, see LangSmith.
FAQ
How is this different from LangGraph or LangChain?
LangGraph is the graph runtime. LangChain's create_agent is a minimal agent harness on top of it. Deep Agents is a more opinionated harness on top of create_agent — same building blocks, but with filesystem, sub-agents, context management, and skills bundled in. For how the three relate, see the LangChain ecosystem overview.
Does this work with open-weight or local models?
Yes. Any model that supports tool calling works — frontier APIs (OpenAI, Anthropic, Google), open-weight models hosted on providers like Baseten or Fireworks, and self-hosted models via Ollama, vLLM, or llama.cpp. Use any LangChain chat model.
Can I use this in production?
Yes! Deep Agents is built on LangGraph, designed for production agent deployments. Pair it with LangSmith for tracing, evaluation, and monitoring. See Going to production for the full guide.
When should I use Deep Agents vs. LangChain or LangGraph directly?
All three are layers in the same stack — see the LangChain ecosystem overview for how they relate. Use Deep Agents when you want the full harness — planning, context management, delegation — out of the box. Use LangChain's create_agent when you want a lighter harness without the bundled middleware. Drop to LangGraph when the agent loop itself isn't the right shape and you need a custom graph.
The layers compose: any LangGraph CompiledStateGraph can be passed in as a sub-agent to a Deep Agent, so custom orchestration plugs in alongside the harness's defaults.
Resources
- Examples — working agents and patterns
- Documentation — conceptual overviews and guides
- LangChain ecosystem overview — how Deep Agents, LangChain, LangGraph, and LangSmith fit together
- API reference — complete reference for all public classes, functions, and types
- Discussions — community forum for technical questions, ideas, and feedback
- LangChain Academy — Comprehensive, free courses on LangChain libraries and products, made by the LangChain team.
- Contributing Guide — how to contribute and find good first issues
- Code of Conduct — community guidelines and standards
Acknowledgements
Inspired by Claude Code: an attempt to identify what makes it general-purpose, and push that further.
Security
Deep Agents follows a "trust the LLM" model. The agent can do anything its tools allow. Enforce boundaries at the tool/sandbox level, not by expecting the model to self-police. See the security policy for more information.
Install the whole package (17 skills):
npx skills add https://github.com/langchain-ai/deepagentsOr install a single skill:
npx skills add https://github.com/langchain-ai/deepagents --skill <name>Pick the skill name from the Skills tab — each entry there installs independently.