When developing with Jazz in watch mode, editing schema.ts now automatically reloads the browser to pick up your changes. Previously you had to manually refresh the page after pushing a schema update to see it take effect. I added a callback that fires after a successful schema push and tells the Vite or SvelteKit dev server to send a full-reload signal to the browser.
On Expo (React Native), Jazz uses a custom fetch wrapper that bridges web-style network calls to the native layer. The web fetch standard allows passing a URL or Request object as input, but the native bridge only accepts plain strings — so libraries like better-auth that use this pattern were crashing. I fixed it by normalising those inputs to a plain string URL plus options before passing them to the native bridge.
When you change your database schema (a "migration"), existing data should stay visible under the new schema. There was a bug where rows that existed before a migration would silently disappear — the resolved table name wasn't being carried through to the query engine, causing a "table not found" error that just dropped the rows silently. I fixed this by making sure the resolved table name is passed all the way through the schema-aware query path.
I added content-based deduplication to the permissions bundle publishing step. When you deploy without changing your schema's access rules, the system now detects that the new bundle is identical to the current one and skips creating a new version — preventing unnecessary version bumps.
I updated Jazz's Svelte query API to accept getter functions — functions that return a query — in addition to plain query objects. This means reactive query updates now flow through automatically, matching what was already possible in the Vue integration.
I updated the Jazz command-line tool to recognise environment variables with framework-specific prefixes — like VITE_, NEXT_PUBLIC_, PUBLIC_, and EXPO_PUBLIC_. Previously, if you set your app ID using your framework's conventional prefix, the CLI wouldn't find it; now it checks all common prefixes automatically.
I extended our automated tests to cover three starter templates that previously had no test coverage. The tests check that the scaffolding tool creates the right files and structure, so we can catch regressions before they reach developers.
I reorganised Jazz's "recipes" (short how-to guides) into clearer sub-folders — access control, auth, and data patterns — so they're easier to find. I also added a new recipe showing how to configure group-level permissions for shared data.
I fixed a race condition in the React chat example where the message input would become usable before the server had confirmed the user was a member of the chat room. I added a readiness check to prevent this, and wrote a Playwright end-to-end test to guard against it happening again.
I fixed a bug where Jazz's SvelteKit plugin didn't apply its configuration correctly the very first time you started the dev server. The plugin now restarts Vite once after allocating the app ID, so SvelteKit's server hooks pick up the config properly.
I fixed broken configuration and authentication in several of Jazz's example applications. The affected apps — including WorkOS auth, Better Auth, Next.js CSR/SSR, Cloudflare Workers, and managed runtime — had issues that stopped developers from using them as working references.
I removed leftover jazz_principal_id JWT claims from the starter templates. Jazz only reads the sub claim to identify users, so this extra claim wasn't doing anything — it was just confusing people who were reading the starter code and thinking it was important.
I restructured the auth documentation to be organized around mental models with framework-specific tabs. I also moved the more technical theory into a new internals reference page covering things like Ed25519 key derivation, JWT structure, and how provider upgrades are verified.
I updated the create-jazz CLI spinner to show progress counts while it fetches GitHub dependencies, so you can see how far along it is instead of staring at a static spinner. I also cleaned up some console warnings that were corrupting the output.
I added jazz-wasm as a direct dependency to six Vite-based starter templates. When using pnpm, packages only get hoisted to the top level if they're listed as direct dependencies, so without this change the starters would fail to resolve jazz-wasm at runtime.
I fixed a bug where JazzProvider in React would tear down and recreate the Jazz client unnecessarily. The issue was that a config object in the useEffect dependency array was triggering cleanup on every render. I switched it to use a stable string key instead, which prevents the cycle.
I updated the Jazz Vite plugins to automatically inject worker.format: "es" and optimizeDeps.exclude: ["jazz-wasm"] via a config hook. This removes the need for developers to add these settings manually to their Vite config.
I added instructions to the project's CLAUDE.md telling AI coding agents to investigate unexpected test failures rather than assuming they're pre-existing, and to never rewrite tests without asking the developer first.
I replaced two placeholder texts in the "What is Jazz?" blog post with proper markdown links — one pointing to the Discord community and one to the booking calendar.
I migrated several example apps (chat-react, moon-lander-react, wequencer) to use the new jazzPlugin() from jazz-tools/dev, switched wequencer to local-first auth, fixed a WebSocket proxy issue, and cleaned up permission configurations.