One page per feature
The reasons, the flow, the data, the contract and the tests belong together. When they live on one page, a change in one is seen next to the others.
PrimeDoc started from a familiar frustration. To understand one feature you opened a document, a drawing, a schema tool, Swagger and Postman, and no two of them agreed. Each was right once. None of them said which one to trust now.
So we built the page we wanted: the reasons, the flow, the tables, the API contract and the tests together, kept current because everyone works on the same one. Backend and frontend developers, testers, product people and AI agents all read and change the same page, each with the access they were given.
PrimeDoc is in early access. It is used for real features today, and it changes quickly with what the people using it tell us.
The reasons, the flow, the data, the contract and the tests belong together. When they live on one page, a change in one is seen next to the others.
The API definition sits next to the words that explain it, the mock server answers from it, and the tests run against it.
Nothing is open until someone shares it, and a link never tells a person without access more than it should.
AI agents can read a project and suggest changes. An editor reviews and applies every one of them.
Pages, diagrams, schemas, contracts and collections export in open formats at any time, and PrimeDoc can run on your own server.
PostgreSQL, plain HTTP and open formats (Markdown, Mermaid, DBML, OpenAPI, Postman), so what you write outlives any one tool.