Developers
The Core API surface.
Core gives your application a persistent personal layer that belongs to the user. Here is what that involves, and how far along each part is.
Core API
In developmentRead and write the parts of a Core a user has granted you.
Authentication
In developmentThe user signs in to their Prifina account as part of connecting.
Permissions and scopes
In developmentAsk for what you need, get only what the user approves.
MCP
AvailableConnect Core to MCP capable clients and coding agents.
Webhooks and events
Architecturally supportedHear about changes in a Core you are connected to.
SDKs
Architecturally supportedClient libraries for common runtimes.
Manifests
Architecturally supportedDeclare what your app is and what it wants access to.
Agent access
Architecturally supportedRun against agent definitions the user keeps in their Core.
Provenance model
In developmentEvery write carries its source, so the trail stays intact.
Write back
In developmentReturn what was learned or decided so it is available elsewhere.
Portability
Architecturally supportedSame integration against any compatible Core, not only ours.
Shape of an integration
01
Register your app
Name it and declare the scopes you need.
02
Add Connect
One action in your interface, no signup form.
03
Read what you were given
Context, preferences, agents or the project at hand.
04
Write back what you produce
Attributed to your app, kept with its source.
Written for humans and for coding agents
The documentation is being built so a coding agent can read it and produce a working integration without a person filling in the gaps. Hosted Core is available to subscribe now.