Statehouse Documentation
Welcome to the Statehouse documentation.
Statehouse is a strongly consistent state and memory engine designed for AI agents, workflows, and automation systems. It provides durable, versioned, replayable state with clear semantics.
Quick Links
- What is Statehouse? - Overview and core concepts
- Getting Started - Installation and first steps
- Python SDK - Python client library documentation
- Examples - Real-world usage examples
Documentation Structure
Introduction
Learn about Statehouse's purpose, design philosophy, and when to use it (or not).
Concepts
Deep dive into core concepts: state and events, transactions, versions, replay, and determinism.
Getting Started
Step-by-step guides to install Statehouse, run the daemon, and perform your first operations.
Python SDK
Complete reference for the Python client library, including transactions, reads, replay, and error handling.
gRPC Internals
For advanced users: understand the gRPC protocol, API versioning, and streaming replay.
Operations
Production guidance: configuration, data directories, snapshots, recovery, and observability.
Failure Modes
What happens when things go wrong: process crashes, disk full, partial writes, and transaction timeouts.
Examples
Real-world examples including the reference agent, crash recovery, and audit workflows.
Status: This documentation covers the MVP release. Some features may be marked as "Draft / MVP" where appropriate.