Skip to content

fix(website): label untitled sessions in the grid picker, Sessions page and Older Sessions - #22987

Merged
KiroCrewCaptain merged 1 commit into
mainfrom
fix/22049-untitled-labels-grid-sessions-history
Oct 22, 2026
Merged

KiroCrewCaptain merged 1 commit into
mainfrom
fix/22049-untitled-labels-grid-sessions-history

Conversation

@iamwhatever

Copy link
Copy Markdown
Contributor

Problem / Motivation

Goal: An untitled session shows "Untitled session" instead of its internal slot key in the session grid picker, on the Sessions page and in the chat sidebar's Older Sessions rows.

Three surfaces fell back to title || key. A session with no title, or with its key as its title, showed a machine id such as chat-9-1785440411 as its name.

Why it matters

The key means nothing to a reader. Two untitled rows looked like two random ids, and the Older Sessions hover tooltip repeated the id.

Not a goal

  • The other title || key sites listed in Route the remaining untitled-session labels through sessionDisplayTitle #22049 (session tabs, chat pane, command palette providers, composer dock, notices, system rows, memory tab). They follow in later PRs.
  • Changing search, sort, navigation, resume or delete. Each surface keeps its keys.
  • The Sessions page avatar letter, which reads the agent name first.
  • New locale strings.

What changed (motivation → approach → change)

Surface Before After
Session grid picker row (components/SessionGridView.tsx) key Untitled session
Sessions page row (pages/SessionsPage.tsx) key Untitled session
Older Sessions row text and hover title (pages/chat-sidebar/HistoryPane.tsx) key Untitled session

All three read the shared sessionDisplayTitle helper from #22041. It treats a missing title and a title equal to the key as untitled. A real title shows as before, and a title that arrives later replaces the label on the next render.

Only the shown label changes. The grid picker still finds a session by its key and opens that key. A Sessions page click still goes to /chat?sid=<key>. An Older Sessions row still resumes its key, and the resumeFromHistory payload title stays title || key.

The history.md spec section on the Sessions sidebar states the rule for the three surfaces.

Backwards compatibility

Compatible: only the displayed label of an untitled session changes; keys, routes, payloads and stored data are untouched.

Removes nothing: no path, key, record or function is removed; search reads the same fields as before.

Tests

  • website/src/test/SessionGridViewCoverage.test.tsx: an untitled row and a key-titled row read "Untitled session" and not their keys; a named row keeps its name; typing the key still finds the untitled row and picking it opens that key. Replaces the case that pinned the key fallback.
  • website/src/test/SessionsPage.test.tsx: the same three rows on the Sessions page; a click on the untitled row navigates with its key.
  • website/src/test/ChatSidebar.historyUntitledLabel.test.tsx (new): renders the real sidebar and opens Older Sessions; untitled rows read "Untitled session" in text and hover title; Enter on one resumes with (key, key), the unchanged payload.

All three cases fail with the production files reverted to origin/main and pass with them. Determinism: the three files passed 3 of 3 runs with --sequence.shuffle.

Manual verification

Rendered the real SessionGridView, SessionsPage and ChatSidebar on a loopback vite server with stubbed API answers, in dark and light themes. Before each capture the script asserted that no untitled key appears in the rendered text.

Screenshots / video

Session grid picker, light:

grid picker light

Sessions page, light:

sessions page light

Older Sessions, dark:

older sessions dark

Other theme

grid picker dark

sessions page dark

older sessions light

Related Issues

Part of #22049

Closes #20873
Closes #20875
Closes #20876

Pattern harvest

Rule candidate: semgrep
Pattern: a session label rendered as title || key in website/src shows an internal id; route it through sessionDisplayTitle.

Checklist

  • At most two commits (one is the norm), with a Conventional Commits title
  • Existing tests pass and new tests added for new functionality
  • Self-review completed; code follows project style guidelines
  • Documentation updated (if applicable)
  • No secrets, credentials, or internal references in the diff

@iamwhatever
iamwhatever requested review from a team and KiroCrewCaptain as code owners October 22, 2026 01:39
@iamwhatever
iamwhatever requested a review from CrysisDeu October 22, 2026 01:39
@github-actions github-actions Bot added the readiness: checking Automated validation is still running label Oct 22, 2026
@github-actions

github-actions Bot commented Oct 22, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Opus 5.5) — ✅ PASS

Premise-level review of d748a30bdd1b91de9e6d018f2b3ca54e4a183465 — why this exists and whether the shipped surface is the smallest honest version. Updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

First-Principles-Verdict: PASS

Nothing blocks. Each of the three surfaces reuses the existing sessionDisplayTitle helper. The remaining title || key sites (11 matches in 7 files outside tests) are deferred in ## Not a goal and tracked in #22049.

What this change ships

Inventory (6 items) — 6 justified

Intent: a person sees "Untitled session" instead of a machine slot id when a session has no real title. This is a FIX (#20873, #20875, #20876).

  1. Untitled rows in the session grid picker read "Untitled session" instead of the key — justified
  2. Untitled rows on the Sessions page read "Untitled session" instead of the key — justified
  3. Untitled Older Sessions rows read "Untitled session" in their text — justified
  4. The Older Sessions hover tooltip reads "Untitled session" instead of repeating the key — justified
  5. A session whose title equals its key now counts as untitled on these three surfaces — justified
  6. The test that pinned the grid picker's key fallback is replaced, which the description states and the linked issues back — justified

[FIRST-PRINCIPLES-REVIEWED] d748a30

Description read

[DESCRIPTION-READ] 1325a49ffa3008bf7f2f5903e7506bb3864fb5aee3ffeb92f03277c2a7a2ff3c

That sha256 names the PR description this verdict read. Recompute it with .github/scripts/pr-description-capture.sh; if it differs, the description changed after this verdict was formed and any finding drawn from the description is unproven.

@github-actions

github-actions Bot commented Oct 22, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Opus 5.5) — ✅ PASS

Design-level review of d748a30bdd1b91de9e6d018f2b3ca54e4a183465 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

Design-Verdict: PASS

Nothing to check. Three label call sites now use the shared sessionDisplayTitle helper. Keys, routes and the resume payload don't change, and the description includes screenshots of all three surfaces.

Notes

  • Several untitled rows on one surface now share the same label, so the only way to tell them apart is the agent, the date or the message count. If users complain, the fix is a secondary hint (such as the date). It doesn't block this PR.

[DESIGN-REVIEWED] d748a30

Description read

[DESCRIPTION-READ] aa2c018bb5ead558fc5662a813e5d2eef637921c0fe371e85cfeaca8c4e68811

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 .github/scripts/pr-description-capture.sh over the same inputs; if it differs, the description or that evidence changed after this verdict was formed and any finding drawn from either is unproven.

@github-actions

github-actions Bot commented Oct 22, 2026 •

Copy link
Copy Markdown
Contributor

Opus 5.5 Review — ✅ no blocking findings

Reviewed d748a30bdd1b91de9e6d018f2b3ca54e4a183465 — this comment is updated in place on each push.

Review details

No findings.

[OPUS-REVIEWED] d748a30

Verdict parsed from the review's SHA-scoped output markers for commit d748a30bdd1b91de9e6d018f2b3ca54e4a183465.

False positive or not applicable? A repository writer can comment:
/ai-review override fable d748a30bdd1b91de9e6d018f2b3ca54e4a183465: <one-sentence reason>

@github-actions

github-actions Bot commented Oct 22, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Opus 5.5) — ✅ PASS

UX-level review of d748a30bdd1b91de9e6d018f2b3ca54e4a183465 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

UX-Verdict: PASS

Nothing blocks. The blind reader read "Untitled session" as "conversations that were never given a name" on all three surfaces, in both themes.

Notes

  • When two rows both say "Untitled session", the reader "can't tell the two apart". In the grid picker, only "4 msgs" and "2 msgs" separate them. A follow-up could show a secondary line, such as the last-active time, next to the label on untitled rows.

[UX-REVIEWED] d748a30

Description read

[DESCRIPTION-READ] 68bf1174c029ad151d4ab6a7989b77df24b4a4177225b750ef357971fbbdde9b

That sha256 names everything this verdict read: the PR description, plus the 9 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 .github/scripts/pr-description-capture.sh over the same inputs; if it differs, the description or that evidence changed after this verdict was formed and any finding drawn from either is unproven.

@github-actions

github-actions Bot commented Oct 22, 2026 •

Copy link
Copy Markdown
Contributor

GPT 6.1 Review — ✅ no blocking findings

GPT 6.1 completed its review of d748a30bdd1b91de9e6d018f2b3ca54e4a183465 and found no blocking issues.

This comment is updated in place on each push.

Review details

No findings.
[GPT-REVIEWED] d748a30

False positive or not applicable? A repository writer can comment:
/ai-review override gpt d748a30bdd1b91de9e6d018f2b3ca54e4a183465: <one-sentence reason>

@iamwhatever

Copy link
Copy Markdown
Contributor Author

Real Adapter Contract Tests is red here for a reason outside this PR. The same three tests in test/test_pi_real_adapter.py fail on main at this PR's base (01327a3, run 39704958002) with the same SandboxUnavailableError. It is tracked in #22959 and #22958. This PR changes only frontend labels, so I am not rerunning or changing anything for this lane; it should clear once the fix for #22959 lands on main.

🤖 AI-assisted comment 🤖

@iamwhatever
iamwhatever force-pushed the fix/22049-untitled-labels-grid-sessions-history branch from 9504393 to d748a30 Compare October 22, 2026 02:24
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Rebased onto main (01327a3 -> 3d6cda4) to pick up #22951, which clears both reds this head showed: Real Adapter Contract Tests (#22959) and the card ratchet in Backend Tests (#22968). Main's same commit failed both; this PR touches neither. New head d748a30; CI re-running.

🤖 AI-assisted comment 🤖

@github-actions github-actions Bot added readiness: passed Eligible automated validation passed for the current revision and removed readiness: checking Automated validation is still running labels Oct 22, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Review-ready on head d748a30: all 105 checks pass, readiness: passed, mergeable. GPT, Opus, Design, First Principles and UX all pass with no findings, so there are no dispositions to answer.

Main has moved one commit since this head went green, in docs/system-specs/modules/app-notifications.md only. That is the same folder as this PR's history.md edit, not the same file, and there is no conflict, so I am not rebasing again.

This PR fixes the grid picker, the Sessions page and Older Sessions, and closes #20873, #20875 and #20876. #22049 stays open for the remaining sites. It now waits only on a maintainer review. I have not merged, approved or turned on auto-merge.

🤖 AI-assisted comment 🤖

@KiroCrewCaptain
KiroCrewCaptain enabled auto-merge (squash) October 22, 2026 04:05

@KiroCrewCaptain KiroCrewCaptain left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Tier 1 auto-approve: fix (7 files). Criteria: no conflict, no requested changes, security path denylist clean, design-doc gate clean, security checklist all-NO, no reviewer-lane findings, readiness signal fresh for this head SHA, declared category matches the diff. Category: the session grid picker, the Sessions page and Older Sessions show the shared untitled-session label instead of the raw slot key. Spec files changed as a ride-along (a minority of the diff on both file count and changed lines), not reviewed as a design decision: docs/system-specs/modules/history.md.

@KiroCrewCaptain
KiroCrewCaptain merged commit 7a1fd5e into main Oct 22, 2026
105 checks passed
@KiroCrewCaptain
KiroCrewCaptain deleted the fix/22049-untitled-labels-grid-sessions-history branch October 22, 2026 04:05
@github-actions github-actions Bot removed the readiness: passed Eligible automated validation passed for the current revision label Oct 22, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

2 participants