Repository navigation
fix: dock bot dot shows only for unread and clears once read - #35726
Conversation
Design Review (Opus 5.5) — ✅ PASSDesign-level review of Design-Verdict: PASS Nothing blocks: the dock dot now means one thing, and the description's table and spec edit cover the change, including dropping needs-you. Notes
[DESIGN-REVIEWED] 05b211b Description read[DESCRIPTION-READ] 2e005fa2489fef69bcb611babcf776c854171a1326d83ab31a1a3fc2a322cfaa That sha256 names everything this verdict read: the PR description, plus the 1 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with |
First Principles Review (Opus 5.5) — ✅ PASSPremise-level review of First-Principles-Verdict: PASS The reported defect, that working kept the dot lit after reading, is fixed where it starts. Removing the needs-you dot goes further than that defect and has no evidence behind it, but it can be undone, so it is a note. Notes
What this change shipsInventory (4 items) — 3 justifiedIntent: a person wants a starred Bot's dock dot to point at a reply they have not read, and to clear once they read it. FIX.
[FIRST-PRINCIPLES-REVIEWED] 05b211b Description read[DESCRIPTION-READ] 6c0afdf0a8e562914e31834eb74fb375b3b3ca08a0ba32fc6155b2d90bf9d06d That sha256 names the PR description this verdict read. Recompute it with |
UX Review (Opus 5.5) — ✅ PASSUX-level review of UX-Verdict: PASS Nothing blocks. The blind reader took the dot to mean "something new/unread", which is what this PR makes it mean, and the before/after shots cover the change in both rail widths. Notes
[UX-REVIEWED] 05b211b Description read[DESCRIPTION-READ] 62c413ff1afe45700ebb2ec750c177c0d9eda9d62f3f44d4dda4e19da10c5790 That sha256 names everything this verdict read: the PR description, plus the 7 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with |
Opus 5.5 Review — ✅ no blocking findingsReviewed Verdict parsed from the review's SHA-scoped output markers for commit False positive or not applicable? A repository writer can comment: |
GPT 6.1 Review — ✅ no blocking findingsGPT 6.1 completed its review of This comment is updated in place on each push. Review detailsNo findings. False positive or not applicable? A repository writer can comment: |
dadbad6 to
077f6e7
Compare
|
Rebased: main moved in website/src/shell/nav/matesDock.tsx, website/src/shell/nav/matesDock.test.tsx and docs/system-specs/modules (cli.md, cloud.md, instances.md) since this head went green (e1b079c -> 80cf3df); CI re-running. The base now carries #35725 (Coverage Gate fix for main) and #35648 (the wide row's dot sits after the name); the conflict kept #35648's in-flow dot class and this PR's unread-only condition. |
Code change approver recommendation: A person should decideWhy: the decider thinks the change is probably fine but estimates a 15% chance that this change introduces a defect, above the 12% limit for approving on its own. MateNavItem now draws the dock dot only when the Bot's DM slot is in unreadSlots, so reading clears it even while the Bot or its worker keeps working. Working still shows on the avatar face for both rail surfaces. The crew-mode spec and the mateNeedsYouByKey comment are updated in the same change. The tests would fail on main: unread-only dot, dot clears after read while still working, no dot for working or needs-you. Dropping the needs-you dot is an explicit, documented part of the stated goal ("only when unread"), and no docs/decisions entry records the dock dot behaviour. The Bots page rows are unchanged, as the non-goals say. CI is green except the rate-limited readiness publisher. An automated recommendation, not a review: a maintainer still decides. It approves on its own only when the estimated chance of a defect is at most 12% and the change has at most 400 changed lines. Decider v3c (217fe607) · commit |
077f6e7 to
05b211b
Compare
|
Rebased: main moved in docs/system-specs/modules (messaging.md) since this head went red on a main-owned test (80cf3df -> 54b06e2); CI re-running. The base now carries #35746, which fixes the TestAmiArchitectureCheck red (#35745) that failed Backend Tests (3.12, 2) and Backend Tests (Windows) (2) on 077f6e7. |
bolichen97
left a comment
There was a problem hiding this comment.
Read the whole diff, tests and the spec edit included, and the linked issue.
MateNavItemnow reads onlyunreadSlotsfor the dock dot. Working stays on the avatar face through theworkingprop, as before, and a turn or dispatched session that waits on the person no longer adds a dot. This is the behavior the issue asks for in its "What it should do" section, so the dot clears the moment the thread is read.- The dot keeps its accessible name, now the fixed unread label, and its position classes for the touch, collapsed and expanded rails are unchanged.
mateNeedsYouByKeyis no longer imported by the dock, and its doc comment is updated to say only the Mates page rows read it. - The tests cover the face for a working mate on both surfaces, the dot only for unread and cleared on read even while a worker runs, and no dot for a turn waiting on the person. The crew-mode spec paragraph matches the new rule.
- The PR description has before and after screenshots. No regex added.
I did not run the tests locally.
Problem / Motivation
Goal: A starred Bot's dot in the left dock shows only when its chat has an unread reply, and goes away as soon as the person reads it.
The dock dot lit for three reasons: the Bot is working, its chat is unread, or it waits on the person. All three drew the same dot. So while a Bot (or a worker it sent) kept working, the dot stayed on even after the person read the reply. On nightly this read as "reading does not clear the dot".
Why it matters
The dot is the dock's notification mark. If it stays on after reading, people stop trusting it and miss the replies it should point to.
Not a goal
What changed (motivation → approach → change)
The dot in
matesDock.tsx(MateNavItem) now reads one fact: is the Bot's chat inunreadSlots. Reading the chat removes it fromunreadSlots, so the dot goes at once.Working still shows on the avatar's face, as before. The
crew-mode.mdspec and themateNeedsYouByKeycomment increwMode.tsare updated to match.Backwards compatibility
Compatible: only when the dock dot appears changes; no data, API or setting is touched.
Reader: website/src/shell/nav/matesDock.tsx:MateNavItem -- crew page -- the working signal it no longer draws as a dot is still drawn on the CrewAvatar face (
face-adata-working assertions in matesDock.test.tsx)Reader: website/src/pages/members/MembersPageRoster.tsx:mateNeedsYouByKey -- crew page -- still the needs-you source for the Bots page rows, unchanged
Tests
matesDock.test.tsxnow locks in: the dot shows only for an unread reply; it drops once read even while the Bot still works; working shows on the face with no dot; a turn or dispatched session waiting on the person adds no dot.Proof: not provable -- frontend test (prove.py runs Python only)
Manual verification
Built the dashboard and shot the dock on main and on this branch with the same seeded data: Unflaky working, Issue Triage unread, Docs Keeper waiting on you. Before, all three had a dot. After, only Issue Triage does.
Screenshots / video
Before (main):
After:
Narrow dock
Related Issues
Closes #34851
Sibling sweep
Swept: git grep -nE "unreadSlots|busySlotKeys" -- website/src/shell
Sibling: website/src/pages/members/CrewmateRowBody.tsx -- not affected -- the Bots page row shows working and needs-you on purpose
Sibling: website/src/shell/nav/matesDock.tsx -- changed here
Pattern harvest
Not generalizable: one dot was fed by three signals by design choice; this is a product rule change, not a code pattern.
Checklist
feat|fix|docs|style|refactor|perf|test|chore|ci|build|revert: ...)