Documentation

Everything about nuthatch - the two-minute golden path, the concepts underneath, and full guides for building nests and operating them. New here? Start with the Quickstart.

Start here

• Quickstart - from a contract address to a live, queryable API in under two minutes.
• What is a nest? - the core mental model.

Build a nest

• nuthatch.toml - the one config file a nest has.
• Authored SQL views - first-class derived logic.
• Recipes - the eth_call you don't need.

Operate

• Run it in production - start here: the whole path from a fresh box to a nest serving unattended, ending in a pre-flight checklist.
• Deploy it - systemd, Docker, a proxy in front, and backups.
• Security - read before exposing /sql: the advisories, and what the guards actually stop.
• Performance - what it measures at, and the three things that decide your backfill's wall clock.
• Run many nests - one runtime, one or more chains, tenancy included.
• Scaled mode - a fleet across machines, when one box is no longer enough.
• The nest registry - publish and pull nests by name.
• Upgrading a nest - the N-1 problem, solved.
• Troubleshooting - starting with "is it me or the endpoint?".
• Verifying a deployment - prove it works on your own hardware, and see what we have and have not verified ourselves.

Reference & internals

• CLI reference, HTTP API, MCP.
• Internals - how it actually works, component by component: the ingest loop, the decode registry, the two storage layers, reorgs, the IVM core, and what changes in scaled mode.
• RFCs - the numbered design record.