LangChain / LangGraph
First met in aicp-c3-basecamp-02 · Dev libraries & frameworks
LangChain / LangGraphFirst met in aicp-c3-basecamp-02 · Dev libraries & frameworks
The code layer under most agent tooling — LangChain for chains and tools, LangGraph for stateful multi-step agents.
You're writing the agent in code rather than on a canvas — or you want to know what n8n's AI nodes are actually doing.
n8n's AI features are built on it, per the BC2 trainer — the canvas is a UI over this
LangGraph handles genuinely stateful agents, not just linear chains
The course's advice: point a coding agent at the docs rather than learning it by hand
Provides the building blocks for agents: models, memory, tools, chains
Runs stateful multi-step agent graphs via LangGraph
Doesn't suit non-technical builders directly — the recommendation is to let a coding agent write against it
Doesn't replace a visual platform for workflows a team has to maintain
Underneath everything agentic. Worth knowing exists even if you never import it.
Dev libraries & frameworks
No agent code in the pipeline.
Only relevant if the vector phase grows into something agentic rather than a search index.
current
n8n: n8n's AI layer is built on LangChain
claude-code: the recommended way to use it — have the agent write the code
| Session | Coverage | Moment | What happened |
|---|---|---|---|
| Basecamp 2: Vibe Coding | explained | 0:35:59 | Code layer under n8n-style tools; agents via LangGraph; non-technical builders should point coding agents at the docs instead of hand-building |
| The Five-Level AI Generalist Roadmap (+ Operating System & Monetization) | mentioned | L2 orchestration and L4 agent frameworks (with LlamaIndex, CrewAI, AutoGen) | |
| AI Sprint: OpenClaw — Session 1 (Infrastructure, Not Agent: the Seven Documents, 1-Click VPS, and Model Routing) | mentioned | Framed as an agentic framework — a different layer than a runtime | |
| AI Sprint: AI Evals & Reliability — Day 2 (Testing Systems That Never Answer Twice) | mentioned | Orchestration references; LangGraph suggested for storing and rating dataset/golden records |