langchain-architecture

Design LLM applications using LangChain 1.x and LangGraph for agents, memory, and tool integration. Use when building LangChain applications, implementing AI agents, or creating complex LLM workflows.

2.34x

Quality

66%

Does it follow best practices?

Impact

82%

2.34x

Average score across 3 eval scenarios

Securityby

Risky

Do not use without reviewing

Optimize this skill with Tessl

npx tessl skill review --optimize ./plugins/llm-application-dev/skills/langchain-architecture/SKILL.md

Evaluation results

97%

93%

Customer Support AI Assistant

ReAct agent with structured tools and session memory

Criteria

Without context

With context

create_react_agent used

100%

Correct LLM class

100%

Recommended model name

100%

MemorySaver checkpointer

100%

Checkpointer passed to agent

100%

Thread ID for sessions

100%

Async invocation

100%

Pydantic schema for complex tool

70%

Safe math eval

100%

Separate sessions demonstrated

40%

100%

LangGraph package used

100%

84%

54%

Internal Knowledge Base Q&A System

RAG pipeline with StateGraph and embeddings

Criteria

Without context

With context

StateGraph used

100%

TypedDict state

100%

START/END constants

57%

100%

VoyageAI embeddings class

100%

voyage-3-large model

100%

PineconeVectorStore used

100%

Retriever k=4

100%

Conditional edges for branching

100%

Async node functions

ChatPromptTemplate used

100%

LangSmith env vars

Claude model used

100%

65%

-8%

AI-Powered Content Production Pipeline

Multi-agent supervisor orchestration with streaming

Criteria

Without context

With context

Supervisor node pattern

100%

Agents return to supervisor

100%

Conditional routing from supervisor

100%

asyncio.gather for batch

100%

astream_events used

100%

astream_events version v2

100%

LangSmith env vars

62%

create_react_agent for specialists

100%

TypedDict state with messages

37%

100%

langchain-anthropic for LLM

LangChain 1.x imports

100%

83%

Async throughout

50%

100%

Repository: wshobson/agents
Commit: 91fe43e

Evaluated: 27 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Customer Support AI Assistant Internal Knowledge Base Q&A System AI-Powered Content Production Pipeline

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.