fix: dedupe Windows notification quick replies (#3467)

[jeanfbrito]

Closes #3467

What

Windows notification quick-reply was sending the message twice on 4.16.0 / dev. This forward-ports the fix from #3420 (b62b165 on hotfix/4.15.6) onto dev with the same semantics.

Thanks to @Moku151 for reporting the 4.16.0 regression in #3467 and for opening #3468. That PR correctly identified the missing forward-port; this PR is Jean’s own port of #3420 rather than a merge of #3468 (third-party forward-port of the same code).

Why

The 4.15.6 hotfix landed the repliedNotifications Set guard, but it was not forward-ported to dev, so 4.16.0 shipped without it. Electron 42 on Windows can emit reply twice for one toast (WinRT + COM activation paths).

How

Same semantics as #3420:

• Guard NOTIFICATIONS_NOTIFICATION_REPLIED with a per-notification repliedNotifications Set
• Re-arm (delete the id) on show
• Never clear on close — the duplicate can arrive after dismiss; clearing there reopens the race (ignore CodeRabbit advice to clear on close)
• Regression specs in src/notifications/main/setup.main.spec.ts: duplicate reply dispatches once; later show then reply is accepted; close does not re-arm
• AGENTS.md: documents the reply-dedup constraint (and that a future handleActivation path must use the same Set) plus the hotfix → dev forward-port rule

Out of scope: CORE-2571, SUP-1097, and the broader Action Center / handleActivation work in #3464.

Testing

yarn test --testPathPattern='src/notifications/main/setup.main.spec'

8 passed (including the three new reply-dedup cases).

  

Summary by CodeRabbit

• Bug Fixes

  • Prevented duplicate notification reply events from being processed.
  • Reply handling now resets when a notification is shown again, while closing a notification does not re-enable duplicate replies.

• Tests

  • Added coverage for repeated replies, notification re-display, and close behavior.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: f9b48172-460a-4bc5-a19c-d87758d3ad2d

📥 Commits

Reviewing files that changed from the base of the PR and between 150b36c and d1ec293.

📒 Files selected for processing (3)

• AGENTS.md
• src/notifications/main.ts
• src/notifications/main/setup.main.spec.ts

Included review availability: Your plan provides up to 4 included reviews per hour; 3 remain after this review.

📜 Recent review details

⏰ Context from checks skipped due to timeout. (3)

• GitHub Check: check (ubuntu-latest)
• GitHub Check: check (macos-latest)
• GitHub Check: check (windows-latest)

🧰 Additional context used

📓 Path-based instructions (6)

**/*.{ts,tsx}

📄 CodeRabbit inference engine (AGENTS.md)

**/*.{ts,tsx}: TypeScript strict mode.

• Redux actions follow FSA (Flux Standard Action) shape.
• No unnecessary comments — self-documenting code through clear naming.
• Verify work with the narrowest meaningful checks first, then broader checks
  when risk or shared behavior justifies it.
  Prefer optional chaining and fallbacks for platform-specific APIs:

Files:

• src/notifications/main.ts
• src/notifications/main/setup.main.spec.ts

**/*

📄 CodeRabbit inference engine (AGENTS.md)

**/*: - File naming: camelCase for files, PascalCase for components.

• Avoid subjective descriptors ("smart", "excellent", "dumb").
• Never invent metrics — no estimated time spent, no speculated user counts.
  Only include numbers from actual logs, error messages, or documented
  sources.

Files:

• src/notifications/main.ts
• AGENTS.md
• src/notifications/main/setup.main.spec.ts

AGENTS.md

📄 CodeRabbit inference engine (CLAUDE.md)

AGENTS.md: AGENTS.md is the canonical project guide, imported
above.

Files:

• AGENTS.md

{AGENTS.md,CLAUDE.md}

📄 CodeRabbit inference engine (CLAUDE.md)

{AGENTS.md,CLAUDE.md}: Route by content, not by request phrasing: shared project guidance,
conventions, or memory ("add this to CLAUDE.md", # memory notes, /init
output) belongs in AGENTS.md — do NOT append it here even if the request
names this file. Claude-only mechanics (skill/hook behavior, Claude-specific
config that no other agent uses) stay below this line, in this file.

Files:

• AGENTS.md

**/*.spec.ts

📄 CodeRabbit inference engine (AGENTS.md)

• Renderer specs use *.spec.ts / *.spec.tsx.

Files:

• src/notifications/main/setup.main.spec.ts

**/*.main.spec.ts

📄 CodeRabbit inference engine (AGENTS.md)

• Main-process specs use *.main.spec.ts.

Files:

• src/notifications/main/setup.main.spec.ts

🧠 Learnings (1)

📓 Common learnings

Learnt from: CR
Repo: RocketChat/Rocket.Chat.Electron

Timestamp: 2026-08-24T02:58:56.120Z
Learning: - Never commit or push without explicit user permission — "fix this" does NOT
  mean "commit it".

Learnt from: CR
Repo: RocketChat/Rocket.Chat.Electron

Timestamp: 2026-08-24T02:58:56.120Z
Learning: - Never commit directly to `master`, `dev`, or `release/X.Y.x` — create a
  branch, test, open a PR.

🔇 Additional comments (3)

AGENTS.md (1)

39-41: LGTM!

Also applies to: 140-148

src/notifications/main.ts (1)

51-51: LGTM!

Also applies to: 82-84, 132-137

src/notifications/main/setup.main.spec.ts (1)

237-299: LGTM!

---

Walkthrough

The notification process now deduplicates reply events per notification. Showing a notification re-arms reply handling, while closing it does not. Tests cover duplicate replies and both lifecycle behaviors. Hotfix guidance now requires immediate or same-change forward-porting to dev.

Changes

Notification reply deduplication

Layer / File(s)	Summary
Reply event guard src/notifications/main.ts, AGENTS.md	A repliedNotifications set suppresses duplicate reply events. The notification ID is removed on show and retained on close.
Reply guard validation src/notifications/main/setup.main.spec.ts	Tests verify one dispatch for duplicate replies, re-arming after show, and suppression after close.

Hotfix forward-port guidance

Layer / File(s)	Summary
Forward-port instruction AGENTS.md	Hotfix changes must be cherry-picked to dev in the same change or immediately after.

Estimated code review effort: 2 (Simple) | ~10 minutes

Merge Risk: ⚪ Minimal · up to d1ec2

This change prevents duplicate Windows notification replies while preserving re-arming behavior for later notifications. No actionable merge-blocking risk remains beyond normal checks and review.

Suggested labels: type: bug

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly describes the Windows notification quick-reply deduplication fix and identifies the related issue.
Linked Issues check	✅ Passed	The changes satisfy #3467 by deduplicating reply events, re-arming on show, retaining the guard after close, and adding regression tests.
Out of Scope Changes check	✅ Passed	The code, tests, and AGENTS.md guidance directly support the linked issue and stated deduplication objective.
Docstring Coverage	✅ Passed	Docstring check was indeterminate for this PR — some files could not be analyzed in time. Not blocking.

Warning

Some tools did not complete. Review the errors below.

🔧 ESLint

If the error stems from missing dependencies, add them to the package.json file. For unrecoverable errors (e.g., due to private dependencies), disable the tool in the CodeRabbit configuration.

src/notifications/main.ts

ESLint skipped: missing config or dependency (missing-dependency). The ESLint configuration references a package that is not available in the sandbox.

src/notifications/main/setup.main.spec.ts

ESLint skipped: the matched ESLint configuration already failed (missing-dependency).

Warning

Errors were encountered while retrieving linked issues.

Errors (2)

• CORE-2571: Request failed with status code 401
• SUP-1097: Request failed with status code 401

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[Moku151]

Thanks for crediting me. Just out of curiosity, could you clarify why a separate
PR was preferred over adding the additional test and documentation to #3468?
I’d like to understand the preferred contribution workflow.

[jeanfbrito]

Thanks for crediting me. Just out of curiosity, could you clarify why a separate PR was preferred over adding the additional test and documentation to #3468? I’d like to understand the preferred contribution workflow.

It was my mistake, I saw your PR in the night and asked the bot to review it and do the fixes needed. He then understood that were to create a new PR. Already fixed, your colaboration will stay forever with us. Thank you very much @Moku151 .