What problem does it solve? Building LangChain agents often leads to outdated patterns, missing memory, runaway loops, and incorrect result handling. This Skill provides the current canonical patterns for creating agents with create_agent, defining tools, adding middleware, and avoiding common mistakes. ## Core Features & Use Cases - Agent Creation: Build agents with create_agent using model strings or instances, system prompts, tools, and checkpointers in both Python and TypeScript. - Middleware & Human-in-the-Loop: Add HumanInTheLoopMiddleware for approval workflows and custom hooks via wrap_tool_call or createMiddleware. - Structured Output & Memory: Get typed Pydantic or Zod responses with response_format, and persist conversation state with MemorySaver and thread_id. - Use Case: You need an agent that searches the web, remembers prior conversation turns, and asks for approval before running a dangerous tool. This Skill shows the exact create_agent configuration, checkpointer setup, and interrupt/resume flow. ## Quick Start Ask the AI to create a LangChain agent with create_agent that uses a custom tool, a MemorySaver checkpointer, and a recursion limit.
