Repository navigation
feat: show a file tab's full path as its hover tooltip - #7935
Conversation
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
1 similar comment
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
Open PR relationship auditThis is a consolidated, point-in-time code-level audit note. It compares complete merge-base diffs and current/merged code; it does not treat a shared topic as duplication or partial coverage as completion. Relationship findings
No PR, Issue, label, branch, or review state was changed by the relationship-note portion of this audit. |
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
3 similar comments
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
2 similar comments
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
88db15f to
7fc442b
Compare
|
Rebased onto main Conflicts: none, clean rebase. The single line this PR edits in Gates run locally: Please review the rebased head. Note that a maintainer push makes the maintainer the last pusher, so under this repo's last-push rule a second approver is needed before merge. Reply here if anything looks wrong. |
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
|
Kiro Crew [operator: chenmingwei23#de330d0c]: This PR has been inactive for 7+ days with failing CI. I've assessed the blockers and they appear resolvable -- I'll push fixes directly to this branch as a co-author. Assessment: Two mechanical blockers, no design change. (1) The branch is If you'd prefer I don't touch this PR, add the |
7fc442b to
baff61d
Compare
|
Rebased onto current Conflict resolved ( Merged #9868 landed an unconditional
Tests run (Node 24, Tree verified against the merge base: 2 files, +130/-2 — component plus its test, nothing else. No open questions left undecided on this one. |
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
2 similar comments
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
|
👋 Hi! This PR's description is missing some required sections from our PR template. Workflow runs won't be auto-approved until the description is updated. Missing sections:
Please update your PR description to include these sections, then push or re-save the description. The workflows will be approved on the next cycle. |
GPT 5.6 Review (fork) — ✅ no blocking findingsReviewed Review detailsNo findings. |
Opus 4.8 Review (fork) — ✅ no blocking findingsReviewed |
UX Review (Fable 5, fork) — 🟡 CONCERNSUX-level review of UX-Verdict: CONCERNS Full-path tooltip is shown but no first-time reader has cold-read it — fork-lane cap, not an author gap. Watch
[UX-REVIEWED] 1667aa1 |
Design Review (Fable 5, fork) — ✅ PASSDesign-level review of Design-Verdict: PASS Verify a file tab and its same-path diff tab stay tellable apart on hover, since both tooltips now show the identical path. [DESIGN-REVIEWED] 1667aa1 |
First Principles Review (Fable 5, fork) — ✅ PASSPremise-level review of First-Principles-Verdict: PASS Base already tooltipped truncated names via The description's "carried no What this change shipsInventory (3 items) — 3 justifiedIntent: let a user identify which file a side-panel tab holds by hovering it — a FIX for #7900.
[FIRST-PRINCIPLES-REVIEWED] 1667aa1 |
The shard-4 red is not from this PR
Failing test What differs. The test AST-scans The four handlers still exist and still refuse a switch mid-turn. Only the attribution moved: their inline Introducing commit. The enumeration test itself is newer than that refactor ( Reproduced on pristine main, no PR merge ref, at Tracking / fix. Already filed and already being fixed — nothing is needed from this PR:
I applied 12718 and 12715 to a pristine main worktree in turn; each takes the file to Please ignore the shard-4 / Coverage Gate red here. It clears on this PR once a fix lands on main and the branch is rebased. No push to this PR is required. |
Committed under `website/temp-screenshots/`, which is where this repo's other 19 evidence PNGs live, and attached to the PR description as well so the fork review lanes can open it (they never check out the fork head, so a committed copy reaches them as a binary marker only). The chip's tooltip is a native `title`, which the browser draws as chrome rather than as page content, so it cannot appear in a page screenshot. The frame instead shows the real tab strip from the built bundle with a bubble painted from the chip's own `title` attribute, read out of the DOM at capture time -- the same pattern the existing capture harnesses use for a `title`. Top half is origin/main, bottom half is this branch, same two tabs and the same hovered chip in both: before, hovering either chip answers `index.ts`; after, it answers the full path of the file that chip holds.
ea9f95b to
1667aa1
Compare
|
Visual-evidence pass on What the evidence isThe description now carries the before/after frame as a real description attachment ( Both halves are the REAL built bundle behind the screenshot harnesses' usual gateway-free static server ( One thing to know before reading the frame: a The committed copy is redundant — drop it at merge if you want the author's file setThe same PNG is also committed at The backend red is main's, not this PR's
|
bolichen97
left a comment
There was a problem hiding this comment.
Approved after verified readiness: passed, PR Readiness success on this head, mergeable, no open change requests. [operator: bolichen97 batch review 2026-09-23]
* feat: show a file tab's full path as its hover tooltip A side panel tab chip labels itself with the file's basename inside a 240px truncating span, and carried no tooltip at all unless it was one of the icon-only pinned view chips. So a long filename was unreadable, and two tabs holding a same-named file from different directories were impossible to tell apart without clicking each one. The chip already receives the tab, and file, diff and folder tabs all carry the absolute path on it. Prefer that path as the tooltip, falling back to the title for tabs that have no path, such as terminal and app tabs, and for the icon-only pinned chips whose glyph needs a name. The breadcrumb bar in the panel body already exposes the full path this way; this brings the chip, which is the surface a user hovers first, to the same behaviour. Closes kirodotdev#7900 * docs: add the screenshot evidence for the tab-chip path tooltip Committed under `website/temp-screenshots/`, which is where this repo's other 19 evidence PNGs live, and attached to the PR description as well so the fork review lanes can open it (they never check out the fork head, so a committed copy reaches them as a binary marker only). The chip's tooltip is a native `title`, which the browser draws as chrome rather than as page content, so it cannot appear in a page screenshot. The frame instead shows the real tab strip from the built bundle with a bubble painted from the chip's own `title` attribute, read out of the DOM at capture time -- the same pattern the existing capture harnesses use for a `title`. Top half is origin/main, bottom half is this branch, same two tabs and the same hovered chip in both: before, hovering either chip answers `index.ts`; after, it answers the full path of the file that chip holds. --------- Co-authored-by: Mitraj Parmar <mrpmrp@amazon.com> Co-authored-by: Joe Guo <zejiangg@amazon.com>
Problem / Motivation
A side panel tab chip labels itself with the file's
basenameinside amax-w-[240px]truncating span, and carried notitleattribute at all unless it was one of the icon-only pinned view chips. So on hover the chip tells the user nothing the label did not already say. Two things a user cannot do today:/repo/alpha/index.tsand/repo/beta/index.tsrender the identical chip) without clicking each one and reading the breadcrumb in the panel bodyWhy it matters
Reported in #7900. The truncated label is frequently ambiguous — repos contain many same-named files (
index.ts,config.json,AGENTS.md), and the only way to confirm which one a tab holds is to click it, which changes what the panel is showing just to answer an identity question. The breadcrumb bar in the panel body already carries the full path; the chip is the surface a user hovers first, and it was the one surface that did not.What changed (motivation → approach → change)
Goal: make the chip answer "which file is this?" on hover, without adding a second tooltip mechanism or new user-facing strings.
Approach: the chip already receives the tab object, and file, diff and folder tabs all carry the absolute path on it as
path. Rather than recomputing or threading a new prop, prefer that existing field as the tooltip and fall back to the current value. The alternative considered — a custom hover card on the chip — was rejected as far more surface for a value the nativetitlealready renders, and the chip'stitleis also the accessible name for the icon-only pinned chips, so it has to stay a single string.The change, in
website/src/pages/chat/SidePanel.tsx:TabChip'stabprop type widens fromPick<PanelTab, 'title'> & Partial<Pick<PanelTab, 'kind' | 'sessionId'>>to include'path', so the optional field is visible to the component.title={tab.title}becomestitle={tab.path ?? tab.title}. Tabs with no path (terminal, app, and the icon-only pinned view chips whose glyph needs a name) are untouched by the nullish fallback.That is the whole source delta: 2 lines changed plus a comment, in one component. No i18n catalog changes, because no new string is introduced — the value shown is a path that already exists in state.
Screenshots
Both halves are the REAL built bundle behind the screenshot harnesses' usual
gateway-free static server, same two tabs, same hovered chip: the top half is
origin/main, the bottom half is this branch.A
titleis drawn by the browser as its own chrome, not as page content, so itcannot appear in a page screenshot (
scripts/capture-reveal-label-sweep.mjsrecords the same limitation). The pale bubble in each half is therefore painted
by the capture harness from the value it reads off that chip's real
titleattribute — the house pattern from
scripts/capture-composer-default-agent-marker.mjs— and the harness asserts the attribute's exact value before it writes a frame,
so a stale bundle fails instead of being saved as evidence:
origin/main: both chips'titleisindex.ts…/packages/checkout/src/index.tsand…/packages/catalog/src/index.tsThe same PNG is committed at
website/temp-screenshots/7935-tabchip-path-tooltip/tabchip-path-tooltip.png,and attached to this description so the review lanes can open it.
Backwards compatibility
Compatible: the prop type only widens (an added optional field), and
??changes the rendered tooltip only for tabs that carry apath— every tab that has none keepstab.titlebyte-for-byte, including the pinned chips that depend ontitleas their accessible name.Tests
New
website/src/test/sidePanelTabChipPathTooltip.test.tsx, four cases, each locking in one half of the ratchet:basenametitlevalues appear and differVerified as a real ratchet, not a tautology: with the source change stashed, the first three assertions fail with
expected null, while the pinned-chip guard passes both ways.Gates run locally in a dedicated worktree:
npx tsc -b— cleannpx vitest run(full frontend suite) — 1764 files, 27,706 passed, 1 expected-fail, 2 skippednpm run jscpd— 0 clonesscripts/ci-surface-tests.py --surface backendselects for a frontend-only diff)Two pre-existing failures on this host, both confirmed identical on a pristine
origin/maintree with the change stashed and the new test file moved out, so neither is introduced here:website/electronnode suite: exit 1 with 1552 tests, 1525 pass, 0 fail, 26cancelledByParentintest/gateway-stop.test.jsops_mission_control/tests/test_routes.py,test_sandbox_argv.py,test_acp_client.py,test_acp_runtime.py,test_browser_cli_install.py,test_terminal_handler.py,test_host_isolation_floor.py,auto_improvement/tests/test_dogfood_learnings.py. These look environment-specific (this host clamped pytest to 2 of 12 workers on available memory, which is itself what the worker-budget assertion checks)No Python source is touched, so
isort,flake8andmypyhave nothing in this diff to act on.Related Issues
Reported in #7900.
Note on CI
Opened from a fork, so the three Bedrock-credentialed AI review checks are skipped by design per CONTRIBUTING.