langchain-ai

Lca Langsmith Essentials — Development skill for Claude Code

Development community

LangChain Academy Course on LangSmith Essentials.

How to install Lca Langsmith Essentials

This entry records only its repository, not the path inside it, so there is no exact command to give. Open langchain-ai/lca-langsmith-essentials and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Lca Langsmith Essentials does

LangChain Academy Course on LangSmith Essentials.

Alternatives in Development

Langchain Skills — ⚠️ — This project is in early development 1.2k ★
Claude Code Everyone Course — Claude Code for Everyone - Learn Claude Code in Claude Code 563 ★
Sun Skill — Justin Sun Cognitive Framework - 孙宇晨认知操作系统 21,829 tweets + 288-page autobiography + 155-episode audio course + 246 ★

README

🍕 LangSlice Pizza Agent

The course repo for **LangSmith Essentials**. LangSlice is a fictional pizzeria, and this agent answers questions about the shop and takes pizza orders. Every run traces to LangSmith.

The agent ships with two deliberate bugs. Finding them in the traces is the exercise.

Repo layout

lca-langsmith-essentials/
├── .env.example                # env template: copy to .env and add your keys
├── env_utils.py                # setup checker, invoked by each project's check_setup.py
├── BUGS.md                     # what the two bugs are (spoilers)
├── langgraph-agent/            # the main agent, the one the lessons follow
│   ├── agent.py                # system prompt, the nine tools, the create_agent call
│   ├── database.py             # knowledge base, ingredient catalog, menu, live orders
│   ├── langgraph.json          # points Studio at the agent and serves the chat UI
│   ├── run_agent.py            # terminal chat
│   ├── eval.py                 # runs the agent over a dataset (Module 2)
│   ├── stock_evaluator.py      # the Module 2 evaluator
│   ├── check_setup.py          # runs the root setup checker against this project
│   ├── ui/                     # customer chat UI, served on port 2024
│   └── traces/                 # recorded conversations, plus upload and download scripts
└── claude-sdk-agent/           # the same pizzeria on the Claude Agent SDK, same filenames

Each agent folder keeps its own `pyproject.toml`, `uv.lock`, and `.venv`. There is nothing to install at the repo root.

Two implementations

Project	Harness	Model	Interfaces
`langgraph-agent/`	LangChain `create_agent` + LangGraph	`openai:gpt-5-nano`	Studio, chat UI, terminal
`claude-sdk-agent/`	Claude Agent SDK	`claude-opus-5`	terminal

Same system prompt, same nine tools, same database, same two bugs. The lessons follow `langgraph-agent/`, so every command below runs from in there unless i