langgraph-cli

RUFEN SIE DIESE FÄHIGKEIT auf, wenn Sie die langgraph-CLI verwenden, um LangGraph-Anwendungen zu erstellen, zu entwickeln, zu bauen oder bereitzustellen. Deckt langgraph new, dev, build, up, deploy und… ab.

npx skills add https://github.com/langchain-ai/langchain-skills --skill langgraph-cli
The `langgraph` CLI manages the full lifecycle of LangGraph applications — from scaffolding a new project to deploying it to LangGraph Platform (LangSmith Deployments).

Key commands:

  • langgraph new — Scaffold a project from a template
  • langgraph dev — Run locally with hot reload (no Docker)
  • langgraph build — Build a Docker image
  • langgraph up — Launch locally via Docker Compose
  • langgraph deploy — Ship to LangGraph Platform
  • langgraph dockerfile — Generate a Dockerfile

All commands (except new) read from a langgraph.json config file in the project root.

When to use

Use this skill when the user wants to:

  • Scaffold a new LangGraph project
  • Run a local development or production-like server
  • Build or deploy a LangGraph application
  • Understand or edit langgraph.json configuration
  • Manage LangSmith Deployments (list, delete, view logs)

Installation

# Python
pip install 'langgraph-cli[inmem]'   # includes langgraph dev support
pip install langgraph-cli             # without dev server (build/up/deploy only)

# if using UV as package manager
uv add "langgraph-cli[inmem]"       # includes langgraph dev support
uv add langgraph-cli                # without dev server (build/up/deploy only)

# JavaScript
npx @langchain/langgraph-cli         # use on demand
npm install -g @langchain/langgraph-cli  # install globally (available as langgraphjs)

Commands

langgraph new [PATH]

Scaffold a new project from a template.

langgraph new                          # interactive template selection
langgraph new ./my-agent               # create in specific directory
langgraph new --template agent-python  # skip prompt, use template directly

Available templates: deep-agent-python, deep-agent-js, agent-python, new-langgraph-project-python, new-langgraph-project-js

langgraph dev

Run a local development server with hot reloading. No Docker required.

langgraph dev                              # default: localhost:2024
langgraph dev --port 8000                  # custom port
langgraph dev --config ./langgraph.json    # explicit config path
langgraph dev --no-reload                  # disable hot reload
langgraph dev --no-browser                 # don't auto-open LangGraph Studio
langgraph dev --host 0.0.0.0              # bind to all interfaces (trusted networks only)
langgraph dev --tunnel                     # expose via Cloudflare tunnel for remote access
langgraph dev --debug-port 5678            # enable remote debugger (requires debugpy)
langgraph dev --n-jobs-per-worker 20       # max concurrent jobs per worker (default: 10)

langgraph build

Build a Docker image for the LangGraph API server.

langgraph build -t my-image                # required: tag the image
langgraph build -t my-image --no-pull      # use locally-built base images
langgraph build -t my-image -c langgraph.json  # explicit config
langgraph build -t my-image --base-image langchain/langgraph-server:0.2.18  # pin base version

langgraph up

Launch the LangGraph API server via Docker Compose (includes Postgres).

langgraph up                               # default port 8123
langgraph up --port 8000                   # custom port
langgraph up --watch                       # restart on file changes
langgraph up --recreate                    # force fresh build (useful for pre-deploy validation)
langgraph up --postgres-uri postgresql://...  # external Postgres
langgraph up --no-pull                     # use local images (after langgraph build)
langgraph up --image my-image              # skip build, use pre-built image
langgraph up -d docker-compose.yml         # add extra Docker services
langgraph up --debugger-port 8124          # serve debugger UI
langgraph up --wait                        # block until services are healthy

langgraph deploy

Build and deploy to LangGraph Platform (LangSmith Deployments). Requires Docker. On Apple Silicon (M1/M2/M3), Docker Buildx is also required for cross-compiling to linux/amd64.

langgraph deploy                           # deploy, name defaults to directory name
langgraph deploy --name my-agent           # explicit deployment name
langgraph deploy --deployment-type prod    # production deployment (default: dev)
langgraph deploy --tag v1.2.0              # custom image tag (default: latest)
langgraph deploy --deployment-id <id>      # update an existing deployment by ID
langgraph deploy --config ./langgraph.json # explicit config path
langgraph deploy --no-wait                 # don't wait for deployment status
langgraph deploy --verbose                 # show detailed server logs

Prereq: LANGSMITH_API_KEY in environment or .env.

langgraph deploy also accepts build flags: --base-image, --pull/--no-pull.

langgraph deploy list

langgraph deploy list                      # list all deployments
langgraph deploy list --name-contains bot  # filter by name

langgraph deploy delete

langgraph deploy delete <deployment-id>          # interactive confirmation
langgraph deploy delete <deployment-id> --force  # skip confirmation

langgraph deploy logs

langgraph deploy logs                                  # runtime logs, last 100
langgraph deploy logs --name my-agent                  # by deployment name
langgraph deploy logs --deployment-id <id>             # by deployment ID
langgraph deploy logs --type build                     # build logs instead of runtime
langgraph deploy logs -f                               # follow/stream logs
langgraph deploy logs --level error                    # filter by level (debug|info|warning|error|critical)
langgraph deploy logs -q "timeout"                     # search filter
langgraph deploy logs --limit 500                      # more entries
langgraph deploy logs --start-time 2026-03-08T00:00:00Z  # time range

langgraph dockerfile <SAVE_PATH>

Generate a Dockerfile (and optionally Docker Compose files) without building.

langgraph dockerfile ./Dockerfile                      # generate Dockerfile
langgraph dockerfile ./Dockerfile --add-docker-compose # also generate compose + .env + .dockerignore

langgraph.json reference

The configuration file used by all CLI commands (dev, build, up, deploy). Defaults to langgraph.json in the current directory.

Minimal config (Python)

{
    "dependencies": ["."],
    "graphs": {
        "agent": "./my_agent/agent.py:graph"
    },
    "env": "./.env"
}

Minimal config (JavaScript)

{
    "dependencies": ["."],
    "graphs": {
        "agent": "./src/agent.js:graph"
    },
    "env": "./.env"
}

Full config with all keys

{
    "dependencies": [".", "langchain_openai", "./local_package"],
    "graphs": {
        "agent": "./my_agent/agent.py:graph",
        "retriever": "./my_agent/rag.py:rag_graph"
    },
    "env": "./.env",
    "python_version": "3.12",
    "pip_config_file": "./pip.conf",
    "dockerfile_lines": [
        "RUN apt-get update && apt-get install -y ffmpeg"
    ]
}

Key reference

KeyRequiredDescription
dependenciesYesArray of dependencies. "." looks for local packages via pyproject.toml, setup.py, requirements.txt, or package.json. Can also be paths to subdirectories ("./my_pkg") or package names ("langchain_openai").
graphsYesMapping of graph ID to path. Format: ./path/to/file.py:variable (Python) or ./path/to/file.js:function (JS). The variable must be a CompiledGraph or a function returning one. Multiple graphs supported.
envNoPath to a .env file (string) OR an inline mapping of env var names to values (object). Used by langgraph dev and langgraph up locally. langgraph deploy reads from this file and adds the variables as deployment secrets.
python_versionNo"3.11", "3.12", or "3.13". Defaults to "3.11".
node_versionNoNode.js version for JS projects.
pip_config_fileNoPath to a pip config file for custom package indexes.
dockerfile_linesNoArray of additional Dockerfile lines appended after the base image import. Use for system packages, binaries, or custom setup.

Typical workflow

  1. Scaffoldlanggraph new to create a project from a template.
  2. Configure — Edit langgraph.json: set dependencies, point graphs at your compiled graph(s), add .env.
  3. Developlanggraph dev for rapid local iteration with hot reload (no Docker, port 2024).
  4. Validatelanggraph up --recreate to test in a production-like Docker stack (port 8123, includes Postgres).
  5. Deploylanggraph deploy to ship to LangGraph Platform (LangSmith Deployments).
  6. Monitorlanggraph deploy logs -f to tail runtime logs; --type build for build logs.

langgraph dev vs langgraph up

Featurelanggraph devlanggraph up
Docker requiredNoYes
Installpip install 'langgraph-cli[inmem]'pip install langgraph-cli
Primary useRapid development & testingProduction-like validation
State persistenceIn-memory / pickled to local dirPostgreSQL
Hot reloadingYes (default)Optional (--watch)
Default port20248123
Resource usageLightweightHeavier (Docker containers for server, Postgres, Redis)
IDE debuggingBuilt-in DAP support (--debug-port)Container debugging

Gotchas

  • langgraph deploy requires Docker — On Apple Silicon (M1/M2/M3), Docker Buildx is also required for cross-compiling to linux/amd64.
  • langgraph deploy can only update its own deployments — Deployments created through the LangSmith UI or GitHub integration cannot be updated with langgraph deploy. Use the UI for those.
  • dependencies must include all packages — The dependencies array in langgraph.json must point to where your package config lives (e.g., "." for root). The actual packages are resolved from pyproject.toml, requirements.txt, or package.json at that location.
  • langgraph dev runs without Docker — It runs directly in your environment. If your code depends on system packages (e.g., ffmpeg), they must be installed locally. Use langgraph up to validate Docker builds.
  • JavaScript CLI — Use npx @langchain/langgraph-cli <command> (or langgraphjs if installed globally via npm install -g @langchain/langgraph-cli).
  • API keyLANGSMITH_API_KEY is required for langgraph deploy. For langgraph dev, it is optional — the server runs without it, but you won't get traces in LangSmith. Can also be set via LANGGRAPH_HOST_API_KEY or LANGCHAIN_API_KEY.

Mehr Skills von langchain-ai

langgraph-docs
langchain-ai
We need to translate the given English text into German, preserving the name "langgraph-docs" if it appears. The text is a description of an agent skill. The instruction says: "Translate only the text inside <text>. Do not include the name unless it appears in the source text." The name "langgraph-docs" does not appear in the source text, so we should not include it. Also, do not add labels like "description" etc. Just translate the text. The text: "Access LangGraph documentation to build stateful agents and multi-agent workflows. Fetches official LangGraph Python docs covering state machines, graph-based agent design, and human-in-the-loop patterns Prioritizes relevant documentation by query type: implementation guides for how-to questions, concept pages for theory, tutorials for end-to-end examples, and API references for technical details Automatically selects 2–4 most relevant documentation URLs and retrieves their content to answer..." Note: There is a missing period after "patterns" and before "Prioritizes". Also the last part seems cut off:
official
langgraph-human-in-the-loop
langchain-ai
Pausiere die Graph-Ausführung für menschliche Überprüfung, Genehmigung oder Validierung und setze sie dann mit deren Eingabe fort. Erfordert drei Komponenten: einen Checkpointer (InMemorySaver oder PostgresSaver), eine Thread-ID in der Konfiguration und JSON-serialisierbare Interrupt-Payloads. interrupt(value) pausiert und zeigt Daten an; Command(resume=value) setzt fort und gibt diesen Wert an den pausierten Knoten zurück. Der gesamte Code vor interrupt() wird bei Fortsetzung erneut ausgeführt, daher müssen Seiteneffekte idempotent sein (upsert verwenden, nicht insert). Unterstützt Genehmigungs-Workflows,...
official
web-research
langchain-ai
Verwenden Sie diese Fähigkeit für Anfragen im Zusammenhang mit Web-Recherche; sie bietet einen strukturierten Ansatz zur Durchführung umfassender Web-Recherchen.
official
langchain-oss-primer
langchain-ai
BEGINNE HIER IMMER für jedes LangChain-, Deep Agents- oder LangGraph-Agent-Bauprojekt. Erforderlicher Ausgangspunkt, bevor andere Fähigkeiten ausgewählt oder Code geschrieben wird…
official
skill-creator
langchain-ai
Leitfaden zur Erstellung effektiver Skills, die die Fähigkeiten eines Agenten durch spezialisiertes Wissen, Workflows oder Tool-Integrationen erweitern. Verwenden Sie diesen Skill, wenn der Benutzer…
official
social-media
langchain-ai
Erstellt plattformspezifische Social-Media-Beiträge mit recherchierten Inhalten und generierten Begleitbildern. Unterstützt LinkedIn-Beiträge (1.300 Zeichen mit professionellem Ton) und Twitter/X-Threads (280 Zeichen pro Tweet im 1/🧵-Format). Erfordert die Delegierung der Recherche an einen Unteragenten vor dem Schreiben, gefolgt vom Lesen der Ergebnisse, um Genauigkeit und Relevanz sicherzustellen. Generiert automatisch auffällige Social-Bilder mit dem generate_social_image-Tool mit kräftigen, kontrastreichen Kompositionen, optimiert für kleine...
official
deep-agents-memory
langchain-ai
We need to translate the given English text into German. The text describes a pluggable memory and file backends system for Deep Agents. It mentions four backend types and a FilesystemMiddleware with six file operation tools. The name "deep-agents-memory" is not in the text, so we don't include it. We must preserve technical terms like "StateBackend", "StoreBackend", "FilesystemBackend", "CompositeBackend", "FilesystemMiddleware", and the tool names (ls, read_file, etc.). Also preserve "Deep Agents" as is? It's a product name, so keep it. Translate the rest naturally. Let's translate: "Pluggable memory and file backends for Deep Agents with ephemeral, persistent, and hybrid routing options." -> "Steckbare Speicher- und Datei-Backends für Deep Agents mit flüchtigen, persistenten und hybriden Routing-Optionen." "Four backend types: StateBackend (thread-scoped, ephemeral), StoreBackend (cross-session persistent), Filesystem
official
deep-agents-orchestration
langchain-ai
Orchestriere Unteragenten, plane mehrstufige Aufgaben und fordere menschliche Genehmigung für sensible Vorgänge an. Delegiere Arbeit an spezialisierte Unteragenten über das Aufgabenwerkzeug; benutzerdefinierte Unteragenten unterstützen isolierte Werkzeugsätze und Systemaufforderungen, während der standardmäßige "Allzweck"-Unteragent die Hauptagentenkonfiguration übernimmt. Plane und verfolge komplexe Arbeitsabläufe mit write_todos, organisiere Aufgaben in den Status "ausstehend", "in Bearbeitung" und "abgeschlossen"; erfordert eine thread_id für die Beständigkeit über Aufrufe hinweg. Implementiere...
official