These articles document the engineering behind ZumiLabs products - how we build local-first, on-device software that makes the browser behave more like a private desktop computer. This page explains how they are written, verified, and kept current, so you can judge how much to trust them.
Who writes them
The articles are authored by Amith @ ZumiLabs and the engineering team. They are written from the actual implementation - not commissioned marketing copy - which is why most of them read as walkthroughs of a specific technique rather than product announcements.
How they are verified
Every code example is grounded in the real product source. Before publishing (and whenever we revise), we cross-reference each article against the shipping implementation and correct anything that has drifted. Where an example is simplified for readability, we say so rather than presenting it as copy-paste production code.
Each article carries a “Tested with” note stating the Chrome version it was verified against, and a References & further reading section linking the canonical sources - MDN, the relevant WICG/W3C specification, Chrome documentation, or the library's own docs.
How they are maintained
Browser platform APIs move quickly - Chrome's built-in AI APIs especially. We treat these articles as living documents: we revise them as APIs change, flag experimental features whose availability varies by browser version and hardware, and update the verification date when we re-check a piece. Freshness through revision matters more to us than simply publishing more.
Corrections
Found something inaccurate or out of date? We want to know. Email [email protected] and we will correct it and note the update.