CaseMark rebuilds Mike on top of Case.dev
The whole point of this fork: rip out the upstream plumbing and stand Mike back up on CaseMark's own platform.
CaseMark has done a foundational rewrite. Mike originally assumed a particular cloud stack for its database, file storage, and login. CaseMark has swapped all three for Case.dev-native equivalents: documents now live in Case Vault (CaseMark's hosted document store), login moves to Better Auth (an open-source auth library) with per-user AI provider keys held encrypted, and a homegrown compatibility layer keeps the rest of the app talking to the new database without a full rewrite.
The same wave adds a Demo Mode with usage metering and a "Powered by Case" surface for a hosted trial, plus a security pass: rate limiting on the login endpoint, a hard refusal to boot in production with placeholder secrets, and a migration that drops legacy plaintext API-key columns from the database.
It's not pick-and-mix - the pieces are interlocked and assume Case.dev's APIs.
Spotted something wrong? Or know the PR text has fresher detail than the writeup above?