LangGraph
- Category
- AI and LLMs
- Docs outline
- 14 sections
Official docs outline
docs.langchain.com- LangGraph overview
Low-level orchestration for stateful long-running agents and when to use LangGraph.
- Thinking in LangGraph
Designing agents as nodes, edges, and shared state.
- Graph API concepts
State schemas, reducers, nodes, edges, conditional routing, and Command and Send.
- Using the Graph API
Building, compiling, branching, and running StateGraph workflows.
- Functional API
The entrypoint and task decorators as an alternative to explicit graphs.
- Persistence and checkpointers
Threads, checkpoints, state snapshots, and replay for durable state.
- Memory and stores
Short-term thread memory and long-term cross-thread memory stores.
- Interrupts and human-in-the-loop
Pausing graphs for human input and resuming with Command.
- Streaming
Stream modes for state values, updates, messages, and custom events.
- Workflows and agents
Patterns including prompt chaining, routing, parallelization, orchestrator-workers, and agent loops.
- Subgraphs and multi-agent
Composing graphs as nodes and structuring multi-agent systems.
- Time travel
Replaying and forking from prior checkpoints for debugging.
- Runtime and Pregel
Super-step execution model, channels, and recursion limits.
- Deployment and observability
Deploying LangGraph applications and tracing with LangSmith.
Each section links to the official page. Titles and summaries by Extra Credit; the docs belong to their authors.
