Description
langchain-architecture is an agent skill from the wshobson/agents collection for building LLM applications with LangChain 1.x and LangGraph. LangGraph is the standard framework for agents in this design, and it uses a StateGraph to keep state explicit, supports durable execution, lets a person inspect and approve steps in a human-in-the-loop design, and checkpoints progress across sessions. The skill covers the package structure of LangChain 1.x, memory and conversation state, tool integration, connecting an LLM to external data sources and APIs, and building document processing pipelines. It is written for developers building autonomous agents that use tools, multi-step LLM workflows, reusable application components and production-grade applications. The guidance is about architecture: how the parts fit together, where state lives, and how to keep an agent's behavior predictable. An agent built this way still needs its own testing and evaluation, which the skill does not provide. The skill is installed with the skills CLI from the wshobson/agents repository. Its skills.sh page showed about 12.9K installs and passing results from three security audits at the time of the scan.…
Highlights
- LangChain 1.x package structure
- LangGraph StateGraph for explicit state
- Durable execution and checkpointing
- Human-in-the-loop inspection
- Tool integration and memory
- Installs with the skills CLI
Ratings & Reviews
No ratings yet
Be the first to rate langchain-architecture.
People also visit
langgraph-docs
Fetches and references LangGraph Python documentation to build stateful agents
finetuning-method-selection
Decide whether to fine-tune at all, and route to the right method
multi-reviewer-patterns
Coordinate parallel code reviews across multiple quality dimensions
langchain-architecture FAQ
What is LangGraph used for here?
The standard agent framework in the skill, using StateGraph for explicit state management, durable execution and checkpoints.
Does it cover human approval?
Yes. It covers human-in-the-loop inspection of agent steps.
Does it cover memory and tools?
Yes. Conversation memory, state and tool integration are among the topics.
Does it include testing?
No. The skill covers architecture. Testing and evaluation are separate work.
How do I install it?
Run npx skills add https://github.com/wshobson/agents --skill langchain-architecture.
What security audits does it have?
skills.sh lists three audits: Gen Agent Trust Hub Pass, Socket Pass and Snyk Pass.
What are the best langchain-architecture alternatives?
Popular alternatives to langchain-architecture include langgraph-docs, agent-architecture-audit and ai-engineer. See all langchain-architecture alternatives