fix!: Always consume latest registry in updateRegistry

[FrederikBolding]

Always consume latest registry in updateRegistry in an effort to provide a way to retry failed OTA updates without requiring a registry update.

---

Note

Medium Risk
Changes the event contract between SnapRegistryController and SnapController, which affects when preinstalled Snap auto-updates and block/unblock checks run. Risk is moderate because it can alter update frequency/behavior even when the registry database is unchanged or fetch is skipped.

Overview
SnapController now reacts to a new SnapRegistryController:registryUpdated event instead of stateChange-filtering on database, ensuring #handleRegistryUpdate runs even when the registry fetch is skipped (recently fetched), the signature is unchanged, or an update fails.

SnapRegistryController publishes registryUpdated with a databaseUpdated boolean on all update paths, and tests/tooling are updated accordingly, including a new test that verifies preinstalled Snap OTA updates can be retried when the registry database itself did not update.

^{Reviewed by Cursor Bugbot for commit 621a01a. Bugbot is set up for automated code reviews on this repo. Configure here.}

[codecov]

Codecov Report

❌ Patch coverage is 85.71429% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 98.56%. Comparing base (79e8b90) to head (621a01a).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
...ages/snaps-controllers/src/snaps/SnapController.ts	66.66%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #3948   +/-   ##
=======================================
  Coverage   98.56%   98.56%           
=======================================
  Files         427      427           
  Lines       12343    12345    +2     
  Branches     1939     1939           
=======================================
+ Hits        12166    12168    +2     
  Misses        177      177           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[Mrtenz]

One nit, looks good to me otherwise

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Reviewed by Cursor Bugbot for commit 621a01a. Configure here.}

[cursor]

Infinite loop when registry fetch fails with null database

High Severity

Publishing registryUpdated in the #update() error path creates an infinite loop when the database is null and fetches keep failing. #handleRegistryUpdate() fires, calls SnapRegistryController:get, which calls #getDatabase(), which calls requestUpdate() (because database is still null), which calls #update(), which fails again and publishes registryUpdated again — restarting the cycle. The old stateChange subscription with the ({ database }) => database selector prevented this because a failed fetch doesn't change database. The new unfiltered registryUpdated subscription removes that protection. Additionally, lastUpdated is never set in the error path, so #wasRecentlyFetched() never short-circuits the loop.

Additional Locations (1)

• packages/snaps-controllers/src/snaps/SnapController.ts#L976-L985

 

^{Reviewed by Cursor Bugbot for commit 621a01a. Configure here.}