LangGraph Documentation

Overview

This skill provides access to the official LangGraph Python documentation to answer questions and guide implementation with accurate, up-to-date information.

Workflow

1. Fetch the Documentation Index

Use the WebFetch tool to read the documentation index:

WebFetch with url="https://docs.langchain.com/llms.txt" and prompt="Extract the full list of documentation URLs with their descriptions"

This index provides a structured list of all available LangGraph documentation with descriptions.

2. Select Relevant Documentation

Based on the user's question, identify 2-4 most relevant documentation URLs from the index. Prioritize:

• How-to guides for implementation questions (e.g., "How do I create a checkpointer?")
• Concept pages for understanding questions (e.g., "What is a state graph?")
• Tutorials for end-to-end examples (e.g., "Build a chatbot with LangGraph")
• API reference for detailed API information (e.g., "StateGraph parameters")

3. Fetch Selected Documentation

Use the WebFetch tool to read each selected documentation URL:

WebFetch with url="[selected-doc-url]" and prompt="Extract all content including code examples, explanations, and API details"

Read 2-4 documentation pages to gather comprehensive context.

4. Provide Accurate Guidance

After reading the documentation:

• Answer the user's question with information directly from the documentation
• Include code examples from the documentation when relevant
• Cite specific documentation URLs for reference
• Recommend additional documentation pages if the user needs more context

Tips

• Multiple topics: If the question spans multiple topics (e.g., "agents with persistence"), fetch documentation for each topic
• Code examples: Always include working code examples from the documentation
• Version compatibility: Check if the documentation mentions version-specific behavior
• Related concepts: Suggest related documentation that might be helpful for the user's broader goals