Drop seven unreachable alternatives from the Smart search rules

[BigSimmo]

Summary

• Seven modeExpansionRules patterns in src/lib/smart-search-intent.ts listed a long phrase alongside a shorter phrase that the long one always contains. Alternation tries the shorter branch first and it always wins, so the longer branch could never be reached. Two are in Prescribing, one in Tools, and four in Calculators.
• Each pattern now carries only the reachable phrase. Nothing is added, removed, or retargeted, and no rule changes which queries it fires on.

Mode	Was	Now
Prescribing	medicine that needs regular blood tests|regular blood tests	regular blood tests
Prescribing	medicine for alcohol dependence|alcohol dependence	alcohol dependence
Tools	check medication interactions|medication interactions	medication interactions
Calculators	screen depression severity|depression severity	depression severity
Calculators	measure anxiety symptoms|anxiety symptoms	anxiety symptoms
Calculators	screen hazardous drinking|hazardous drinking	hazardous drinking
Calculators	rate obsessive compulsive symptoms|obsessive compulsive symptoms	obsessive compulsive symptoms

Why this is behaviour-identical: in every pair the surviving phrase is a suffix of the deleted one and is preceded there by a space, so the \b boundaries land in the same places. Any string the old pattern matched, the new one matches. This was checked over 126 framed strings per-pattern (prefixes medicine that needs, check, screen, measure, rate, adjacent-character, hyphenation, case and whitespace variants) with zero differences, and the existing unit tests already assert three of these rules using the long phrasing and pass unchanged.

RAG impact: no retrieval behaviour change — src/lib/smart-search-intent.ts is offline catalogue vocabulary expansion for the Smart modes, not a document retrieval or ranking surface, and this change is a proven no-op within it. No file under src/lib/rag/, no retrieval RPC, no golden fixture, and no ranking contract test is touched.

Verification

• npm run test:focused -- --files src/lib/smart-search-intent.ts — 113 test files, 1225 tests passed
• npm run lint — clean, --max-warnings 0
• npm run typecheck — clean
• npm run format — whole-tree, committed

Verification not run: npm run verify:pr-local — its dry run selects full heavy scope including npm run build, deferred to CI, which runs the same required gates on this head.

UI verification not run: no UI, routing, or styling file changed, so npm run verify:ui does not apply.

npm run eval:retrieval:quality does not apply. This is catalogue term expansion, not document retrieval, ranking, selection, chunking, or scoring.

No tests were added or changed. The equivalence proof is that the existing tests, which use the long phrasing this PR deletes, still pass against the trimmed patterns.

Risk and rollout

• Risk: very low. Seven lines changed, no lines added, and the change is a proven no-op on a pure offline module.
• Rollback: revert the single commit.
• Provider or production effects: None. No provider call, no network access, and no schema change.

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 Clinical KB Database (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

Basis for the seven items above: the diff deletes unreachable regex alternatives in one pure, offline module. It makes no provider call, no network request, and no schema or Supabase change, and it neither reads nor writes document content, source metadata, review status, or credentials. Deployment classification was considered and is unchanged, because the change is behaviour-identical and produces no clinical recommendation.

Notes

• Follow-up to Tighten the movement, avoidance, and couples Smart search rules #2692, which fixed three rules that were genuinely mismatching. These seven were flagged in that PR's notes as dead text rather than a defect and deliberately left out to keep that diff to the behaviour change.

🤖 Generated with Claude Code

https://claude.ai/code/session_012j3M3zd5PkGfN5vJe3BiSU

---

Generated by Claude Code

---

Note

Low Risk
Offline catalogue vocabulary expansion only; seven regex simplifications with no intended change to which queries match or which terms expand.

Overview
Removes seven dead regex alternates in modeExpansionRules inside smart-search-intent.ts. Each pattern previously listed a longer phrase and a shorter substring; matching always hit the shorter branch first, so the longer arm was unreachable.

Prescribing (regular blood tests, alcohol dependence), Tools (medication interactions), and Calculators (four severity/symptom phrases) now use only the surviving substring. Associated terms and which queries trigger expansion are unchanged—this is a readability/cleanup pass, not a retargeting of rules.

^{Reviewed by Cursor Bugbot for commit fef557b. Configure here.}

[coderabbitai]

Important

• 🔍 Trigger review

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

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Team

Run ID: 9471c6c2-45e6-47aa-a88c-523e5fe9432b

---

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.}

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-07T09:31:32.846472Z	fef557b	PR opened

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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[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 ↗︎.

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_5df8c4af-ce19-4dd1-914b-31ce64c07e18)

[BigSimmo]

Consolidated into #2736 at db6376e. The complete original head fef557b is retained as an ancestor, including its code, tests and audit requests. Closing this duplicate review/CI stream under the authorised consolidation; source branch retained. Current verification and remaining merge prerequisites are recorded on #2736. This closure does not claim that the successor is merged or that its required CI is green.