PlugWiki

langgraph-cli Skill for Claude: Scaffold, Run and Deploy LangGraph Apps

Makes Claude use the langgraph CLI to scaffold, run, build and deploy LangGraph apps, and edit langgraph.json.

Installs
9.3K
on skills.sh
Stars
1.3K
on GitHub
Updated
Oct 2026
last repo commit

This skill is published by LangChain in its langchain-skills repo. It teaches Claude the langgraph command line tool: create a project from a template, run a dev server with hot reload, build a Docker image, start a production-like stack, and deploy to LangGraph Platform (LangSmith Deployments). It also explains every key in the langgraph.json config file.

It needs the CLI installed (SKILL.md gives the pip, uv and npm commands). langgraph deploy also needs Docker and a LangSmith API key, and it reads your .env file to create deployment secrets. The skill folder is one file with no scripts. The repo README says the project is in early development.

Good for

  • Developers building agents with LangGraph
  • Teams deploying LangGraph apps to LangSmith Deployments
  • Claude Code users who want the right langgraph command without opening the docs

Install langgraph-cli

Claude Code

  1. 1

    Copy this prompt.

    Install the langgraph-cli skill in Claude Code. Run this command:
    
    pip install 'langgraph-cli[inmem]'
    
    Then check that it loaded and tell me how to use it.
  2. 2

    Paste it into Claude Code. It shows the change and asks before running it.

Requires Python.

Pick one route, not both. The plugin installs every skill under config/skills (22 folders in the repo; the README heading says 21) and has no hooks or MCP servers. The skill drives the langgraph CLI, which it installs with: pip install 'langgraph-cli[inmem]' (includes the dev server), or uv add "langgraph-cli[inmem]", or npx @langchain/langgraph-cli. Docker is needed for build, up and deploy.

From the project README. Source.

What's in the skill

When it loads

INVOKE THIS SKILL when using the langgraph CLI to scaffold, develop, build, or deploy LangGraph applications. Covers langgraph new, dev, build, up, deploy, and langgraph.json configuration.

Sections in SKILL.md

  1. When to use
  2. Installation
  3. Commands
  4. langgraph new [PATH]
  5. langgraph dev
  6. langgraph build
  7. langgraph up
  8. langgraph deploy
  9. langgraph dockerfile <SAVE_PATH>
  10. langgraph.json reference
  11. Minimal config (Python)
  12. Minimal config (JavaScript)
  13. Full config with all keys
  14. Key reference
  15. Typical workflow
  16. langgraph dev vs langgraph up

Key facts

Installs on skills.sh
9,327 (Nov 2, 2026)
Last repo commit
Oct 29, 2026

What it includes

  • Scaffolds a project from one of 5 templates, such as agent-python or deep-agent-js

    Interactive by default. --template skips the prompt. Templates: deep-agent-python, deep-agent-js, agent-python, new-langgraph-project-python, new-langgraph-project-js.

    Runs code · langgraph new

  • Runs a local server with hot reload on port 2024, no Docker

    Options include --port, --no-reload, --no-browser, --host 0.0.0.0 (trusted networks only) and --tunnel, which exposes it through a Cloudflare tunnel.

    Runs code · langgraph dev

  • Builds a Docker image for the LangGraph API server

    -t sets the image tag. Options include --no-pull, -c for the config path and --base-image.

    Runs code · langgraph build

  • Starts the server with Docker Compose and Postgres on port 8123

    Options include --watch, --recreate, --postgres-uri, --image and --wait. SKILL.md compares it with dev in a table.

    Runs code · langgraph up

  • Builds and ships to LangSmith Deployments; also deploy list, delete and logs

    Needs Docker and LANGSMITH_API_KEY. It adds variables from your env file as deployment secrets. deploy delete asks to confirm unless --force is set.

    Delete · langgraph deploy

  • Writes a Dockerfile, and optionally Docker Compose files, without building

    --add-docker-compose also writes a compose file, .env and .dockerignore.

    Write · langgraph dockerfile

  • Explains the 7 config keys, with minimal Python and JavaScript examples

    Keys: dependencies, graphs, env, python_version, node_version, pip_config_file and dockerfile_lines.

    langgraph.json reference

From the official docs

Permissions

Medium risk

It runs CLI commands that can deploy and delete deployments in your LangSmith account and send your .env values as secrets. Keep a key scoped to what you need and check the deploy target first.

What it can do

  • Read your data
  • Create and change things
  • Delete things
  • Read-only mode

Watch out for

  • Deploy ships your project to LangSmith

    langgraph deploy builds a Docker image of the project and deploys it to LangSmith Deployments (dev by default, prod with --deployment-type prod) using LANGSMITH_API_KEY. Check the target deployment and type before you approve it. github.com

  • Deploy sends your env file values as secrets

    SKILL.md says langgraph deploy reads the env file named in langgraph.json and adds those variables as deployment secrets. Check what is in that file first. github.com

  • It can delete deployments

    langgraph deploy delete asks for confirmation unless --force is given. Do not let Claude add --force without checking the deployment ID. github.com

  • Some dev options widen access

    --host 0.0.0.0 binds all interfaces (SKILL.md says trusted networks only) and --tunnel exposes the server through a Cloudflare tunnel. github.com

  • langgraph.json can add Dockerfile lines

    The config key dockerfile_lines appends lines to the Dockerfile; SKILL.md's example runs apt-get install. Review any lines Claude adds. github.com

Technical details
Sign-in and keysLANGSMITH_API_KEY (or LANGGRAPH_HOST_API_KEY or LANGCHAIN_API_KEY) from your environment or .env file.
Scopes
  • Runs the langgraph CLI in your project: scaffolds files, builds Docker images and starts local servers

  • langgraph deploy builds your project and ships it to LangSmith Deployments with your LangSmith API key

  • langgraph deploy delete removes a deployment

  • Installs the CLI with pip, uv or npm

Questions

What does the langgraph-cli skill do?

It teaches Claude the langgraph CLI: create a project, run a dev server, build a Docker image, run a production-like stack, deploy to LangSmith Deployments, and edit langgraph.json.

How do I install it in Claude Code?

Use the install box at the top. The plugin route installs all LangChain skills; the skills CLI route installs this one. You also need the langgraph CLI itself, which the box notes cover.

Does it need an API key?

Only for deploying. langgraph deploy needs LANGSMITH_API_KEY. For langgraph dev it is optional: the server runs without it, but you get no traces in LangSmith.

langgraph dev or langgraph up?

dev runs without Docker on port 2024 with hot reload, keeping state in memory or in a local folder. up runs Docker Compose with Postgres on port 8123 and is meant for production-like checks.

Can it deploy for me?

It can run langgraph deploy if you ask. That needs Docker and your LangSmith key, and it can only update deployments it created itself, not ones made in the LangSmith UI or GitHub integration.

Is it made by LangChain?

Yes. It sits in the langchain-ai/langchain-skills repo, and the plugin manifest names LangChain as author.

Does it work in the Claude app?

The README lists Claude Code, Cursor, Windsurf and Codex. The skill drives a local CLI and Docker, so use a client that can run commands.

Is it safe?

The folder is one text file. The risk is in what it has Claude run: deploys and deletes against your LangSmith account, and an env file sent as deployment secrets. Read each command before you approve it.

Information

updated Nov 2, 2026

Folder
config/skills/langgraph-cli

langchain-ai/langchain-skillsskills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 3
  1. SKILL.md · github.com
  2. langchain-skills README · github.com
  3. plugin.json · github.com

Updated Nov 2, 2026.