[BigSimmo]

User description

Summary

Before: Clinical mode looked different from Work mode. It had boxed headers and result strips, long therapy cards, a split Psychiatry home, separate compare controls on each list, and filters that hid rows. Desktop records filled one long column.

After: Clinical uses Work's header, title band and centred column. Lists are calmer and grouped. Therapy and Differentials rows are compact, and the Avoid / modify warning is still shown in full. Psychiatry has one Home page. There is one shared Compare tray, and filters highlight matches instead of hiding the rest. On desktop, DSM, Therapy and Act records sit beside their list, and a cited source opens beside the answer.

The owner approved this on 12 Oct 2026 from high-fidelity mockups ("go ahead with all your recommendations"). The filter change was approved separately in the Ward Flow ideas.

• Areas touched: App experience, Design system, Personal practice, Knowledge and records
• Item 1: the Clinical header gets glass controls and the pill names the area only. A ModeNav title sits above the tabs, and the column is 64rem from 1024 px. The new src/app/clinical-mode.css holds the styles.
• Item 2: SearchResultsHeaderBand is one plain line. DSM chapters are one grouped hairline card.
• Item 3: compact Therapy and Differentials rows. "Best used for" and "Best fit" move into the record word for word. Therapy Review leaves the tab bar and is linked from Therapy search.
• Item 4: Psychiatry tabs are now Home and Saved. ?page=tools lands on Home. Medicines drops "Lives in Psychiatry". The First Nations entry is kept.
• Item 5: the Differentials tidy-up. The count is never clipped, the Must-check shelf is gone (Emergent still leads), and scenario chips sit in one scrolling row.
• Item 6: a SharedCompareTray for Therapy, Differentials and DSM search.
• Item 7: ClinicalSplit and DesktopSplitList, an Act topic list, and a docked answer source panel on desktop only. Phones are pixel-identical.
• Item 8: recent questions show as a list, the Try line fits a 320 px phone, and the desktop shows one logo.
• Filters highlight instead of hiding, on Differentials, DSM and Therapy. A note reads "N of M highlighted, all rows stay" and has a Clear button.
• Favourites at 320 px with large text: the header goes to two rows and the content wraps.
• The Differentials compare dock now claims its phone band only once something is ticked (compare-presence.ts).

Verification

• Checks run:
  • npm run test:focused (combined branch): Test Files 1 failed | 563 passed (564). The one failure was the pill-weight test's source anchor; it was fixed, and tests/mode-pages-sheet-classes.test.ts now gives Tests 4 passed (4).
  • tsc --noEmit (combined branch): exit 0.
  • eslint --max-warnings=0 and prettier on the changed files: clean.
  • check:organisation: every new file is placed.
  • Design-system adoption: 55 components, 232 roots.
  • Builders' focused browser proof: smoke and desktop split specs passed.
  • Three Differentials, DSM and Tools browser tests failed against the dev server and were not proven to fail on base as well. The full browser suite is left to CI.
• UI verification not run: full verify:ui and npm run build are memory-killed in this container; CI runs the browser suite.

Risk and rollout

• Risk: medium, layout and interaction across Clinical. No clinical wording, sources, urgency order or retrieval changed.
• Rollback: revert the merge commit.
• Provider or production effects: None.

Clinical Governance Preflight

• Source-backed claims still require linked source verification before clinical use
• No patient-identifiable document workflow was introduced or expanded without explicit governance approval
• Supabase target remains PsychSift Production (sjrfecxgysukkwxsowpy)
• Service-role keys and private document access remain server-only
• Demo/synthetic content remains clearly separated from real clinical sources
• Source metadata, review status, and outdated/unknown-source behavior remain conservative
• Deployment classification/TGA SaMD impact was checked when clinical decision-support behavior changed

Presentation only. The Avoid / modify warning keeps its full text at every width. Emergent ordering is unchanged. Filtering now keeps every row visible instead of hiding the non-matching ones, which is the more conservative behaviour. No server, data or source code changed.

Notes

• This PR stays a draft. It is clinical, and only the owner sends it live.
• Known limits:
  • The Therapy list's existing 24-row cap now applies per group.
  • The therapy record's Cautions tile still clips with "Tap for detail", as it did before this PR.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Nv6DEPtQxG5JfAUBa2Q1gc

---

Generated by Claude Code

---

Summary by cubic

Pulls Clinical mode into Work's visual language: the header takes Work's glass controls with a calm title band, Therapy and Differentials rows become compact, one shared Compare tray replaces the separate compare controls, and filters now highlight matching rows instead of hiding the rest. From 1024px a record opens beside its list, and a cited source docks beside the answer.

List and record changes

• Psychiatry's tabs are now Home and Saved; the split Ask/Tools pages merge into Home, and Medicines drops its signposts.
• Therapy and Differentials rows are compact; the Avoid / modify warning shows in full at every width, and "Best used for" and "Best fit" move into the record.
• DSM, Therapy and Act records show their list beside them from 1024px, with the current record marked; phones and upright tablets are unchanged.
• A cited source opens docked beside the answer on desktop, dismisses with Escape or its close button without touching browser history, and reserves answer-column room so it never covers the prose.
• Therapy Review leaves the main tabs and is linked from the Therapy search page; the 24-row cap now applies per group.
• Recent questions render as a list, the Try line fits a phone, and favourites wrap at large text on 320px phones.

Filtering and compare

• Differentials, DSM and Therapy filters highlight matching rows, keep the rest under "Not matching", show an "N of M highlighted" note with Clear, and announce changes through the shared announcer.
• Therapy, Differentials and DSM search share one Compare tray, which claims its dock band only once something is ticked.
• Therapy's tray lists each selection in a sheet with a 48px Remove button; the DSM tray keeps a ticked diagnosis title even outside the current results.
• The Differentials Search tab links the results route directly when a query is carried.
• Presentation only — no clinical wording, sources, urgency order or retrieval changed.

^{Written for commit 4e51281. Summary will update on new commits.}

---

CodeAnt-AI Description

Make clinical references easier to scan without hiding useful results

What Changed

• Filters in Differential, DSM, and Therapy search now put matching rows first while keeping other results in view under a “Not matching” or urgency-specific heading. A count explains how many are highlighted, and Clear removes the highlighting filters.
• DSM, Differential, and Therapy lists now use one Compare tray that names selected records, lets users clear the selection, and opens comparison when enough records are selected. On phones it sits above the search box or in the page footer; on wider screens it floats above the list.
• Therapy search uses compact rows with the full Avoid / modify warning visible. Best-use and fit details are shown in the therapy record, and the Review queue remains one tap away from search.
• On wide screens, DSM and Therapy records show a navigation list beside the record, and the Act page lists topics that jump to their sections. Cited sources open beside the answer instead of covering it; smaller screens keep the existing single-column pages and source drawer.
• Psychiatry now has Home and Saved pages. Home brings together question search, shift tools, sections, builders, and recent activity; old links to the Tools page open Home.
• Clinical headers and result counts use a simpler, Work-style layout. Recent questions on the home page appear as full-width tappable rows instead of chips.

Impact

✅ Non-matching clinical records stay accessible while filtering
✅ Full therapy cautions visible in search
✅ Side-by-side clinical references on wide screens

💡 Usage Guide

Checking Your Pull Request

Every time you make a pull request, our system automatically looks through it. We check for security issues, mistakes in how you're setting up your infrastructure, and common code problems. We do this to make sure your changes are solid and won't cause any trouble later.

Talking to CodeAnt AI

Got a question or need a hand with something in your pull request? You can easily get in touch with CodeAnt AI right here. Just type the following in a comment on your pull request, and replace "Your question here" with whatever you want to ask:

@codeant-ai ask: Your question here

This lets you have a chat with CodeAnt AI about your pull request, making it easier to understand and improve your code.

Example

@codeant-ai ask: Can you suggest a safer alternative to storing this secret?

Preserve Org Learnings with CodeAnt

You can record team preferences so CodeAnt AI applies them in future reviews. Reply directly to the specific CodeAnt AI suggestion (in the same thread) and replace "Your feedback here" with your input:

@codeant-ai: Your feedback here

This helps CodeAnt AI learn and adapt to your team's coding style and standards.

Example

@codeant-ai: Do not flag unused imports.

Retrigger review

Ask CodeAnt AI to review the PR again, by typing:

@codeant-ai: review

Check Your Repository Health

To analyze the health of your code repository, visit our dashboard at https://app.codeant.ai. This tool helps you identify potential issues and areas for improvement in your codebase, ensuring your repository maintains high standards of code health.

[supabase]

This pull request has been ignored for the connected project sjrfecxgysukkwxsowpy because there are no changes detected in supabase directory. You can change this behaviour in Project Integrations Settings ↗︎.

---

Preview Branches by Supabase.
Learn more about Supabase Branching ↗︎.

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: ASSERTIVE
• Plan: Advanced
• Run ID: 06b52dd4-daa7-48a0-beae-61ce62c40ea3

---

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

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.}

[codspeed]

Merging this PR will not alter performance

⚠️ 5 benchmarks spent significant time in system calls

System calls cannot be consistently instrumented, so they are not included in the measure, which understates the real cost. Please switch to the Walltime instrument to accurately measure system calls.

Measurement and system calls

• medication-entities: medicationEntityMatchesInText
• chunking: chunkTextWithOverlap long document
• catalog-search: rankCatalogRecords 400 records
• answer-ranking: rankAnswerEvidence 40 results
• chunking: buildChunks 12 pages

ℹ️ 1 benchmark has no flame graph

The flamegraph could not be generated because too much time was spent in runtime internals, making the visualization not useful for performance analysis.

• medication-entities: medicationEntityMatchesInText

✅ 8 untouched benchmarks

---

_{Comparing claude/clinical-polish-mockups-dme269 (4e51281) with main (5f6b272)}

[codeant-ai]

🏁 CodeAnt Quality Gate Results

Commit: 4e51281f
Scan Time: 2026-10-12 17:22:18 UTC

✅ Overall Status: PASSED

Quality Gate Details

Quality Gate	Status	Details
Secrets	✅ PASSED	0 secrets found
Duplicate Code	✅ PASSED	0.0% duplicated
SAST	✅ PASSED	No security issues
Bugs	✅ PASSED	Rating S: No bugs
IAC	✅ PASSED	No IAC issues

View Full Results

[gitar-bot]

Gitar has auto-approved this PR (configure)

[gitar-bot]

Gitar has auto-approved this PR (configure)

[qodo-code-review]

Code review by qodo was updated up to the latest commit ccaf343

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: ccaf343c3a

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[qodo-code-review]

Qodo Fixer

✅ Committed (2) · ☑ Fixed (2)

Commits pushed directly to this PR — no separate fix PR opened.

Process — 2 fixed

• ☑ Fixed: Resizing leaves an extra Back step
• ☑ Fixed: Back can do nothing after an in-page action

[gitar-bot]

Gitar has auto-approved this PR (configure)

[BigSimmo]

Both Qodo Back-button findings on the docked source panel, and Codex's matching P2, are fixed in 3531247. That commit replaces Qodo's automatic fix c1dbf10, which is reverted in 3a043ce.

Qodo's fix changed the shared Sheet, so every drawer in the app would have behaved differently. It also called replaceState on history from a close effect, which can race the history.back() of a normal dismiss. That is the same kind of race that once undid the DSM search query. It also only removed the marker, so the stranded same-URL entry stayed in history.

The smaller fix is that the docked panel no longer adds a history entry, so Back does not close it on desktop. This deliberately undoes c7bd2e0, which made Back close the panel. A non-modal side panel that the reader can leave through the page cannot own a history entry without stranding one. Escape and the close button still dismiss it, and the phone Sheet keeps Back-to-close. A unit test now checks that opening the panel leaves history.length unchanged.

---

Generated by Claude Code

[gitar-bot]

Gitar has auto-approved this PR (configure)

[qodo-code-review]

Code review by qodo was updated up to the latest commit 3531247

[BigSimmo]

The CodSpeed flag on 3531247 (medicationEntityMatchesInText, -58%) is measurement noise. On the previous head, c1dbf10, the same benchmark showed +13.9%. The only change between those two commits is the desktop source panel's history handling, which touches no medication or search code. CodSpeed also reports "different runtime environments" and unmeasured system-call time for that benchmark. CodSpeed is not part of the required checks, and this PR does not fix it.

---

Generated by Claude Code

[gitar-bot]

Gitar has auto-approved this PR (configure)

[qodo-code-review]

Code review by qodo was updated up to the latest commit 86c1f56

[qodo-code-review]

Code review by qodo was updated up to the latest commit 0add1b2

[gitar-bot]

Gitar has auto-approved this PR (configure)

[qodo-code-review]

No code changes since the last review — review skipped

[gitar-bot]

Code Review ✅ Approved 3 closed / 3 findings

🟡 Medium risk · Reworking clinical filtering, comparison, and record navigation across 86 files could make users overlook or misselect references.

Brings Clinical mode in line with Work's header, compact Therapy and Differentials rows, one shared Compare tray, and desktop side-by-side record and source layouts. Clearing a filter is now announced as the doc comment describes. One open minor finding: switching the source from a sheet to the docked panel while open can leave a dead Back entry in src/components/clinical-dashboard/answer-source-drawer.tsx and src/components/clinical-dashboard/answer-source-docked-panel.tsx, which should be consumed before docking. The DSM Compare tray showing raw slugs for ticks outside current results was closed without a fix in this PR.

Auto-approved: No blocking issues found.
Please see Auto-approve Docs for details on setting custom approval criteria.

✅ 3 closed

✅ Edge Case: Clearing a filter is never announced, though the doc comment says it is

📄 src/components/clinical-dashboard/filter-highlight.tsx:17-19 📄 src/components/clinical-dashboard/filter-highlight.tsx:38-42
The new doc comment says turning a filter on, changing it, "or clearing it" is spoken through the shared announcer. When the filter is cleared, though, active becomes false, so message is "". The effect only calls announce when message is truthy, and announce() also drops empty text. So pressing Clear says nothing, and the Clear button unmounts as well, which leaves a screen-reader user with no confirmation. Fix: announce a clear message when the filter goes from active to inactive, for example by tracking the previous active value in a ref.

Closed: Edge Case: DSM Compare tray shows raw slugs for ticks outside current results

📄 src/components/dsm/dsm-search-page.tsx:550-560
The tray's second line looks up each selected slug in queryResults and falls back to the slug itself. selected is component state seeded from initialIds. It survives a client-side query change on /dsm/search, but queryResults only holds the current query's matches. So if a reader ticks diagnoses and then searches for something else, or arrives with ?ids= that the query doesn't match, the tray reads like "major-depressive-disorder · Panic Disorder". Fix: look the title up in the full DSM summary catalogue, or store {slug, title} when a row is ticked so the name doesn't depend on the current results.

Closed: Edge Case: Sheet→dock swap while open still leaves a dead Back entry

📄 src/components/clinical-dashboard/answer-source-drawer.tsx:61-72 📄 src/components/clinical-dashboard/answer-source-docked-panel.tsx:106-110
The drawer still passes dismissOnBack to the phone/tablet Sheet, and opening it there pushes a ps_sheet_* history entry. If the viewport then crosses 1024px while the source is open (for example, an iPad rotated to landscape), SourceDrawerFrame swaps to the docked panel. The Sheet's cleanup deliberately skips history.back(), and the docked panel now closes through onClose alone, so the pushed entry is never removed. The next Back press lands on that entry and does nothing, which is the stranded-Back problem this commit set out to remove. A fix: when the frame switches from Sheet to docked while open and history.state still carries the sheet's key, consume the entry (e.g. history.back() before docking, or replaceState to drop the key).

Review coverage

🤖 Auto-approval Approved

📋 Rules No rules evaluated

🧪 Functional validation Not enabled · Set up

Options

Auto-apply is off → Gitar will not commit updates to this branch.
Display: compact → Counting what did not apply, without listing it.

Comment with these commands to change the behavior for this request:

Auto-apply	Compact
gitar auto-apply:on	gitar display:verbose

_{Was this helpful? React with 👍 / 👎 | Gitar}