Build on data that
stays connected.
Commons gives your application a local, versioned copy of the data it depends on, every published change as it arrives, and capabilities that shape how the data is used. Full documentation is on the way.
What you can expect
from Commons.
Commons handles the hard parts of shared data, so your code works with a local dataset that receives published changes and keeps its history.
Your application reads from its own copy, fast and available even when the publisher isn’t.
Subscribe once and receive each signed, published change. No polling, no diffing files.
Record the version behind a result and retrieve exactly what the data said at that point.
Commons stores and syncs bytes. Capabilities interpret CSV, JSON, documents, and more.
Add ingestion, validation, views, and search as signed, sandboxed releases.
Give assistants tools to list, read, and search a dataset through MCP.
Choose how
you build.
Work from the command line, read data from your application, react to changes, or give an assistant its own tools. Use one, or combine them in the same project.
Arguments, schemas, and full references will live in the documentation.
Extend Commons
with capabilities.
Each capability fills one of three slots: how data comes in, which records are accepted, and how people and tools use it.
Start from a starter, build, and check it in the same sandbox the App uses, then publish a signed release any dataset can attach.
Run it where
your users are.
A Commons instance can run in a data center or on a laptop, with mobile support planned. Each one can publish, serve, or subscribe within its permissions.
Deploy shared nodes for distribution, laptops for local work, and devices in the field, and let them sync with each other.
Each instance holds its own Store. Roles follow permissions, not device type.
Available now.
Coming next.
Commons is in early access. Here is what you can build with today, and what is on the way.
Available in early access
- Local, versioned Stores with signed sync and retained history
- Per-Store access rules
- Signed capability releases, in WebAssembly or declarative form
- CSV, JSON, API, stream, and Key-Value sources
- Table views, full-text search, queries, and MCP tools
- Starters, reproducible builds, and scratch-dataset previews
Coming soon
- Beckn and DeDi capabilities for open networks
- A working examples repository
- Discovery across the wider network
- Database sources and data-lake connectors
- Windows support
- Automatically derived datasets
Documentation
is on its way.
Installation, interfaces, and full command references will live in the docs. Join the waitlist to hear when they are ready.