Technology
How CIVOLINK systems are built and run. These are working engineering principles, not a published standard and not a product API.
In one sentence
Independent applications, shared presentation — one brand over systems that never borrow each other’s privileges.
How CIVOLINK is built
-
Build against a product, not a monolith
Each product is its own application with its own data and its own release cycle. There is no shared core that everything depends on. A change in one product cannot break another by accident.
-
Independent applications, shared presentation
Products present themselves under one brand identity while remaining separate systems. The corporate website is static content: it holds no product data, connects to no product backend, and offers no shared login.
-
Small, explicit, versioned — when they exist
Where interfaces exist, they are small, explicit and versioned. Public API documentation does not exist yet. When it does, it will live at the planned documentation portal — docs.civolink.co.ke future — which is not linked because it does not exist.
-
Boundaries before features
Security boundaries, data ownership and operational responsibility are defined before features are added. Failure of one system does not take another with it. Access is granted per system, never inherited from the brand.
-
A pattern, not a specification
These principles are observed across current engineering work. They are not a contractual specification, a service-level agreement or a guarantee of uptime. Product-specific documentation will be published per product when each product is ready for it.
-
Status honesty in every interface
Anything built under this brand shows its true state: planned, in development, or live. Badges are re-verified before release. A badge reading “Live†on something that is not live is a defect, not a style choice.
How this site is built
Proposed direction — open decision D22. The statements below are the planned approach for the production site, recorded as a reversible assumption for this prototype.
The corporate website is planned as hand-authored static HTML and CSS: no framework, no database, no accounts, no analytics and no third-party assets. Nginx is the approved web server. Content is edited as files; there is no content-management system at launch. The final production stack and hosting topology remain an open decision.
- Semantic HTML5, one shared stylesheet, minimal JavaScript
- System font stack — no web-font downloads, no font CDNs
- Six approved brand colours only; no third-party images
- No cookies, no trackers, no external requests
- Contact form: local prototype only until privacy terms exist
Where technical writing will live
A documentation and support portal is planned at docs.civolink.co.ke future. It does not exist yet and is not linked. Until it does, technical questions go through the contact page.
No public API is claimed. Illustrative payload examples, sample endpoints and integration sandboxes are deliberately absent from this page — they would imply interfaces that do not exist. They belong with the future documentation portal, when there is something real to document.
Planning an integration?
If you are evaluating an integration or a technical partnership, tell us what you are building and which platform you want to work with.