LangChain - Skill

Purpose

This skill teaches the agent project specific conventions and concise patterns for building AI agent workflows with LangChain in Python. Use this when implementing chatbots, RAG pipelines, tool-enabled agents, multi-agent flows, or deploying runnables.

When an AI assistant should apply this skill

• Editing Python code that imports langchain or langchain_core
• Adding or changing agents, tools, retrievers, memory, or chains
• Writing pipeline glue code for RAG and vector stores
• Creating deployable runnables or LangServe endpoints

Quick start

1. Keep imports explicit and small. Prefer the Runnable interfaces for consistency. 2. Use typed small functions for tools. 3. Attach memory to chains or agents when conversation state is required. 4. Wrap external I/O behind tools to make tests deterministic.

Core concepts and cheat sheet