skills.rest
langchain-ai
langchain-aiOFFICIAL
Explore all skills by this author →
Website𝕏@LangChainGitHub

deepagents-python-quickstart

Scaffolds a minimal local Deep Agent in Python using provider-native web search.

1.2k|90|Updated Jan 22, 2026
Search Tags:#web search#Python#Langchain#AI Agent#Quickstart#Deep Agents#Scaffold
One-click install
npx skills add https://github.com/langchain-ai/langchain-skills --skill deepagents-python-quickstart
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: deepagents-python-quickstart
Source: https://github.com/langchain-ai/langchain-skills/tree/main/config/skills/deepagents-python-quickstart
Command: npx skills add https://github.com/langchain-ai/langchain-skills --skill deepagents-python-quickstart

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires deepagents, python-dotenv.

What problem does it solve?

Setting up a first Deep Agent in Python often leads to stale API guesses, unnecessary third-party search keys like Tavily, and polluted project directories. This Skill walks through the official quickstart with a clean, model-agnostic local setup.

Core Features & Use Cases

  • Official quickstart alignment: Fetches the live LangChain Deep Agents quickstart docs and implements the exact create_deep_agent research-agent shape shown there.
  • Provider-native web search: Replaces Tavily with the chosen provider's built-in search tool (Anthropic, OpenAI, or Google), so only one API key is needed.
  • Isolated setup: Creates a new directory, installs deepagents plus the provider package, stores the key in a gitignored .env, and runs a research example.
  • Use Case: You want to try Deep Agents locally with Claude or GPT without signing up for a separate search API — the Skill scaffolds and runs a working research agent in minutes.

Quick Start

Use the deepagents-python-quickstart skill to build a minimal local Deep Agent in Python with my chosen model provider's built-in web search.

Frequently Asked Questions about deepagents-python-quickstart

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a Deep Agent in Python locally?▼

Follow the official LangChain Deep Agents quickstart, which uses create_deep_agent with a research system prompt and invokes it with a research question. This Skill fetches the live docs and implements that exact shape in a new isolated directory.

How to use Deep Agents without a Tavily API key?▼

Replace the quickstart's Tavily-based internet_search tool with your model provider's built-in web search. Anthropic, OpenAI, and Google all offer native search tools, so only that provider's API key is required.

Which models work with LangChain Deep Agents?▼

Deep Agents are model-agnostic and accept a provider:model string such as anthropic:claude-sonnet-5, openai:gpt-5.5, or google_genai:gemini-3.5-flash. Anthropic, OpenAI, and Google are preferred because they provide built-in web search.

What packages do I need to install for a Deep Agent quickstart?▼

Install the deepagents package, python-dotenv for environment variables, and the LangChain provider package matching your chosen model. You do not need tavily-python when using provider-native search.

Does the Deep Agents quickstart require LangSmith tracing?▼

No, LangSmith tracing is optional and skipped by default in this setup. The only secret required is your model provider's API key stored in a gitignored .env file.

Related Skills

Explore More Software Engineering Skills→
Varnan-TechVarnan-Tech
581

geo-gap-fixer

Audit brand visibility on LLMs and generate a GEO action plan.

Official
Advanced
langchain-ailangchain-ai
1.1k

Deep Agents Core

Create and configure Deep Agents with LangChain and LangGraph.

Official
Advanced
artgaurav16420-ossartgaurav16420-oss

langsmith-fetch

Fetch and analyze recent LangSmith traces to debug LangChain and LangGraph agents.

Community
Intermediate
JosephRobles23JosephRobles23

langgraph-fundamentals

Teach LangGraph state management, nodes, edges, and conditional routing.

Community
Advanced