Lubu-LabsOfficial·7 Agent Skills Included

langchain-agent-skills

Build, test, and deploy LangGraph and Deep Agents projects

Scaffolds LangGraph and Deep Agents projects, then guides multi-agent patterns, state management, error handling, and task planning. Eliminates guesswork around langgraph.json config, retry policies, checkpointers, and LangSmith trace debugging. Includes runnable scripts for validation, deployment, rollback, testing, and A/B evaluation to ship production agents faster.

101|14

↗

npx skills add Lubu-Labs/langchain-agent-skills --all -g -y

Available:

Tells the agent how this repository's skills are structured and which architecture, tooling, and validation rules to follow when building or modifying LangChain ecosystem skills.

All Skills in This Repository (7)

Pure Emerald Level Indicators

📦 In Repo

Lubu-Labs

langgraph-agent-patterns

Build and validate LangGraph multi-agent graphs with supervisor, router, orchestrator-worker, and handoff patterns.

langgraph-error-handling

Automate LangGraph error handling with RetryPolicy, ToolNode, and HITL interrupts.

langsmith-deployment

Deploy LangGraph agents via LangSmith Deployment across cloud, hybrid, and standalone environments.

langgraph-project-setup

Generate LangGraph project scaffolds with configuration and environment templates.

skill-creator

Create modular skills with SKILL.md frontmatter and starter resource directories.

langgraph-state-management

Design LangGraph state schemas, reducers, and persistence with TypedDict and MessagesState.

langsmith-trace-analyzer

Analyze LangSmith traces to generate Markdown or JSON diagnostics reports.

Official

Advanced

Frequently Asked Questions

FAQPage Schema

How to install langchain-agent-skills?▼

Run `npx skills add Lubu-Labs/langchain-agent-skills --all -g -y` in your terminal to install all skills globally.

What does this LangChain skills collection cover?▼

It covers the full LangGraph and Deep Agents lifecycle: project setup, multi-agent patterns, state management, error handling, testing, trace analysis, and production deployment to LangSmith.

Does it work with Claude Code, Codex, and Cursor?▼

Yes. Every skill uses the standard SKILL.md format, which Claude Code, Codex, OpenCode, and Cursor all read natively.

Can it help debug failing LangGraph agents?▼

Yes. The trace analyzer downloads and organizes LangSmith traces, while the error-handling skill classifies failures and applies retry, recovery, or human escalation patterns.

Do I need deep LangGraph experience to use it?▼

No. Each skill includes quick-start commands, decision tables, and ready-to-run scripts, so your coding assistant handles the framework details for you.