langchain-agent-skills
Nine skills for LangChain, LangGraph, LangSmith and Deep Agents: project setup, agent patterns, testing, traces and deployment.
- 106 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The set consists of instructions and examples and runs nothing itself
- Examples send requests to model services and LangSmith with your keys
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/langchain-agent-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The set consists of instructions and examples and runs nothing itself
- Examples send requests to model services and LangSmith with your keys
This is third-party code. Review the repository files before installing.
What it does
The nine-skill set covers the development path on the LangChain stack: initialising LangGraph projects (structure, `langgraph.json`, environment variables), multi-agent coordination patterns (supervisor, router, orchestrator with workers, handoffs), state schemas, reducers, checkpoints, retry strategies and human escalation. Separate skills test agents with trajectory evaluation and LangSmith datasets, analyse LangSmith traces, help deploy applications in cloud, hybrid and standalone modes, and set up Deep Agents projects with `write_todos` planning. The service skill `skill-creator` describes how to extend the collection. Skills are packaged into three plugins: `langgraph-skills`, `langsmith-skills` and `deepagents-skills`.
Who it is for. Agent developers on LangChain and LangGraph who work with Claude Code or Codex and want ready instructions instead of searching documentation.
Good fit when
- A new LangGraph project must be set up with a correct `langgraph.json` and dependencies.
- Several agents are linked with a supervisor or router pattern.
- An agent fails on external calls and needs retries with human escalation.
- LangSmith traces are analysed to find the cause of slow behaviour.
- Deployment in cloud, hybrid or standalone mode needs a plan and monitoring.
Not a fit when
- The project does not use the LangChain ecosystem.
- The official LangChain vendor skills are needed: this set was assembled by a community.
- You need work without Python or JavaScript: the examples are written in those languages.
Example request
Initialise a LangGraph project with a supervisor and two workers and prepare trajectory tests.Limitations
This is a community set of instructions, examples and helper scripts rather than a LangChain product, so track library versions in their own documentation. Examples are written for Python and JavaScript and need model-provider and LangSmith keys where they call those services. Skills trigger by task description, but can also be pointed to by file path. The repository was last updated on 20 October 2026 and uses the MIT licence.
How to disable. Run `/plugin uninstall` for the installed bundle or delete the skill folders from the agent skills directory. Projects built from them stay in place.
FAQ
How many skills are in the set?
Nine working skills and the service `skill-creator` for extending the collection.
Which plugins is the set split into?
Three plugins: `langgraph-skills`, `langsmith-skills` and `deepagents-skills`; each includes `skill-creator`.
Does the set work in Codex?
Yes. The `SKILL.md` format is identical, and Codex reads skills from `.agents/skills` in a repository and `~/.agents/skills` globally.
What does the LangSmith part cover?
Trace analysis for debugging and optimisation, plus deployment and monitoring of LangGraph apps in cloud, hybrid and standalone modes.
Is this the official LangChain set?
No, it is community-maintained. The vendor official skills are described in a neighbouring card.
Discussion
No questions or reviews yet.
Related
SkillsOfficial LangChain skills: choose between LangChain, LangGraph and Deep Agents, quickstarts, RAG, memory, human approval and eval engineering
A LangGraph 1.0 guide as a skill: ReAct agents, StateGraph, middleware, memory, MCP and RAG, so Claude Code writes LangChain code without outdated APIs
LangSmith skills
LangSmith Skills
Official LangChain skills for LangSmith: query and export traces, build eval datasets and write custom evaluators from your agent
An agent-first CLI: LangSmith projects, traces, runs, datasets and experiments from Claude Code or Cursor.
A catalog of nearly two hundred skills: code delivery, review, debugging, AI governance, product work and agent coordination.
A command-line skill manager: installs skills from a marketplace, GitHub, private Git and npm into 45 agents' folders and scans them first
README in short
The README lists skills in a table, shows plugin installation for Claude Code and copying for Codex, and `AGENTS.md` describes commands for creating and validating new skills.