[TimeToBuildBob]

This adds an integration test for Datastore::close racing active writers.

aw-tauri closes the server's shared datastore at shutdown while tray and settings-API clones may still be writing. ActivityWatch/aw-tauri#302 relied on a one-off probe for this (ten databases, eight writers, 80 assertions). That probe only ever lived in /tmp, so a future change to the worker's Close/commit path would have no regression signal.

aw-datastore/tests/close_durability.rs, run for 3 rounds × 4 writers on a real SQLite file in a tempdir:

• Each writer gets one acknowledged write, then all four race set_key_value against close().
• close() must return within 5 s, so active writers cannot starve it.
• An independent open_reader connection must see each writer's last acknowledged value.
• The closed worker must reject later requests.

Limit: Close does not drain in-flight HTTP handlers (noted on the aw-tauri#302 review). This test covers requests to the worker only.

Verification

• Passes 10/10 locally, about 0.1 s per run.
• Mutation check: with the worker's final commit on Close swapped for a rollback, the test fails because the acknowledged key is missing after Close. Reverted.
• cargo fmt and cargo clippy --test close_durability clean. Test-only change, no production code touched.

[greptile-apps]

Confidence Score: 5/5

[Low impact] The PR appears safe to merge; the earlier unbounded writer-wait issue is fixed.

Summary

Adds a file-backed test that races four writers against Datastore::close across three rounds.

• The datastore suite checks that Close keeps acknowledged writes.

_{Reviews (2) · Last reviewed commit: "test(datastore): bound writer completion..." · Reviewed by Greptile}

[codecov]

Codecov Report

❌ Patch coverage is 96.66667% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.77%. Comparing base (656f3c9) to head (d7217b8).
⚠️ Report is 225 commits behind head on master.

Files with missing lines	Patch %	Lines
aw-datastore/tests/close_durability.rs	96.66%	2 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           master     #840       +/-   ##
===========================================
+ Coverage   70.81%   85.77%   +14.95%     
===========================================
  Files          51       90       +39     
  Lines        2916    12926    +10010     
===========================================
+ Hits         2065    11087     +9022     
- Misses        851     1839      +988     

☔ View full report in Codecov by Harness.
📢 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.

[TimeToBuildBob]

@greptileai review

[TimeToBuildBob]

CI-green and mergeable (Greptile 5/5) — waiting only on a maintainer click.

This PR is ready to merge, but the bot has pull-only access to this repo and can't self-merge — surfacing it here so it isn't lost. The monitoring loop will stop re-flagging it now that this note is posted.

[TimeToBuildBob]

🤖 AI code review

Adds an integration test that races four cloned Datastore writers against close() across three temporary SQLite databases. The test bounds completion time, reads each writer's last acknowledged value through open_reader(), and checks that the closed datastore rejects subsequent requests.

Safe to merge — no P0/P1 findings

Confidence 5/5

✅ No findings. The diff looks correct to me on this pass.

Files changed (1) — the diff as I read it

• aw-datastore/tests/close_durability.rs — Adds close_keeps_every_acknowledged_write_from_racing_writers with synchronized writers, completion timeouts, persisted-value assertions, and a post-close request rejection assertion.

_{Reviewed d7217b822fcf · panel: codex:gpt-6.1-sol · subscription engine · 24s · about this reviewer}

_{Maintainer commands}

_{@TimeToBuildBob review (own line) — fresh review · @TimeToBuildBob fix — a worker acts on the findings. Once per comment; 👀 = received.}

Outage coverage: 1-pass, not merge evidence. Subscription review supplied findings while full consensus capacity was unavailable. The consensus merge gate remains unsatisfied.