Synap documentation
Synap is an open-source data pod: a server that holds your records and lets the AI agents you connect work on them. By default an agent's creates and edits apply directly and can be undone, while deletions and structural changes wait for your approval.
It is for people and small teams who work with AI agents and want their notes, contacts, tasks and documents in one place they own. For example, you ask Claude Code to log a call with a client: it adds the person and a note about the call to your pod, and you find both in the Synap app on your desktop.
Get a pod
Not sure which one? Choose your path.
What do you want to do?
The Browser app on the desktop; Relay for your phone and Raycast are in beta. Synap Cloud or your own pod.
Give Claude Code, Codex, Cursor, ChatGPT or any MCP client its own governed access to your pod.
Shape a pod to your work, from installing a template to writing your own app.
Install, configure, domains and TLS, backups, upgrades and operations.
The SDK for apps people sign in to, API keys and the Hub REST API for scripts, and webhooks.
How Synap works
All concepts, in reading order- 01What is Synap?A data pod you own, with a governed place for AI agents to work on your records.
- 02Kind and facetsOne entity, one kind, many roles: what a thing is, apart from the roles it plays.
- 03ViewsA named query plus a rendering configuration over the same records.
- 04Validation policyWhether a requested change runs directly, becomes a proposal, or is refused.
Words you do not know are in the glossary.