Audit fixes + Node 24 toolchain bump

package.json

@@ -2,7 +2,7 @@
   "name": "prompt-for-codex-medical-knowledge-base",
   "version": "0.1.0",
   "private": true,
-  "packageManager": "npm@10.9.8",
+  "packageManager": "npm@11.12.1",
   "engines": {
     "node": "24.x",
     "npm": "11.x"

src/app/api/documents/[id]/signed-url/route.ts

@@ -1,14 +1,16 @@
 import { NextResponse } from "next/server";
+import { z } from "zod";
 import { getDemoDocument } from "@/lib/demo-data";
 import { env } from "@/lib/env";
 import { isDemoMode } from "@/lib/env";
-import { jsonError } from "@/lib/http";
+import { jsonError, PublicApiError } from "@/lib/http";
 import { createAdminClient } from "@/lib/supabase/admin";
 import { AuthenticationError, requireAuthenticatedUser, unauthorizedResponse } from "@/lib/supabase/auth";
 
 export const runtime = "nodejs";
 
 const signedUrlTtlSeconds = 60 * 10;
+const routeIdSchema = z.string().uuid();
 
 export async function GET(_request: Request, { params }: { params: Promise<{ id: string }> }) {
   try {
@@ -24,6 +26,8 @@ export async function GET(_request: Request, { params }: { params: Promise<{ id:
       });
     }
 
+    if (!routeIdSchema.safeParse(id).success) throw new PublicApiError("Invalid document id.");
+
     const supabase = createAdminClient();
     const user = await requireAuthenticatedUser(_request, supabase);
     const { data: document, error } = await supabase

src/app/api/eval-cases/route.ts

@@ -2,8 +2,9 @@ import { NextResponse } from "next/server";
 import { z } from "zod";
 import { normalizedClinicalSearchTokens } from "@/lib/clinical-search";
 import { clinicalQueryModeSchema } from "@/lib/clinical-query-mode";
-import { isDemoMode } from "@/lib/env";
+import { env, isDemoMode } from "@/lib/env";
 import { jsonError, PublicApiError } from "@/lib/http";
+import { queryPrivacyMetadata, queryTextForStorage } from "@/lib/query-privacy";
 import { searchScopeFiltersSchema } from "@/lib/search-scope";
 import { createAdminClient } from "@/lib/supabase/admin";
 import { AuthenticationError, requireAuthenticatedUser, unauthorizedResponse } from "@/lib/supabase/auth";
@@ -88,7 +89,10 @@ export async function POST(request: Request) {
       .from("rag_query_misses")
       .insert({
         owner_id: user.id,
-        query: parsed.data.query,
+        // Raw clinical queries are potential PHI: store the normalized form unless
+        // raw retention is explicitly enabled (RET-H4), matching every other
+        // rag_query_misses / rag_queries writer.
+        query: queryTextForStorage(parsed.data.query),
         normalized_query: normalizedQuery,
         query_class: parsed.data.queryClass ?? parsed.data.queryMode,
         top_files: sourceFiles,
@@ -101,11 +105,14 @@ export async function POST(request: Request) {
         promoted_eval_case: true,
         promoted_at: new Date().toISOString(),
         metadata: {
+          ...queryPrivacyMetadata(parsed.data.query),
           interaction: "answer_eval_capture",
           rating,
           feedback_type: parsed.data.feedbackType ?? null,
-          note: parsed.data.note,
-          answer: parsed.data.answer,
+          // Verbatim clinical note/answer are the larger PHI surface; only retain
+          // them when raw retention is explicitly enabled (RET-H4).
+          note: env.RAG_PERSIST_RAW_QUERY_TEXT ? parsed.data.note : "",
+          answer: env.RAG_PERSIST_RAW_QUERY_TEXT ? parsed.data.answer : "",
           query_class: parsed.data.queryClass ?? null,
           query_mode: parsed.data.queryMode,
           filters: parsed.data.filters ?? {},

src/app/api/images/[id]/signed-url/route.ts

@@ -1,14 +1,16 @@
 import { NextResponse } from "next/server";
+import { z } from "zod";
 import { getDemoImage } from "@/lib/demo-data";
 import { env } from "@/lib/env";
 import { isDemoMode } from "@/lib/env";
-import { jsonError } from "@/lib/http";
+import { jsonError, PublicApiError } from "@/lib/http";
 import { createAdminClient } from "@/lib/supabase/admin";
 import { AuthenticationError, requireAuthenticatedUser, unauthorizedResponse } from "@/lib/supabase/auth";
 
 export const runtime = "nodejs";
 
 const signedUrlTtlSeconds = 60 * 10;
+const routeIdSchema = z.string().uuid();
 
 export async function GET(_request: Request, { params }: { params: Promise<{ id: string }> }) {
   try {
@@ -25,6 +27,8 @@ export async function GET(_request: Request, { params }: { params: Promise<{ id:
       });
     }
 
+    if (!routeIdSchema.safeParse(id).success) throw new PublicApiError("Invalid image id.");
+
     const supabase = createAdminClient();
     const user = await requireAuthenticatedUser(_request, supabase);
     const { data: image, error } = await supabase

src/app/globals.css

@@ -14,6 +14,7 @@
  *   rounded-xl  bottom sheets, dialogs, large shells
  */
 @theme {
+  --radius-xs: 0.25rem;
   --radius-sm: 0.375rem;
   --radius-md: 0.5rem;
   --radius-lg: 0.625rem;
@@ -152,10 +153,9 @@
   --disabled: #94a3b8;
   --focus: #1aa99b;
 
-  /* Motion primitives (shared by both themes) */
-  --duration-fast: 150ms;
-  --duration-base: 200ms;
-  --duration-slow: 250ms;
+  /* Motion primitives (shared by both themes). The duration values live with
+     the easing tokens lower in this block (120/180/240ms); they are not
+     redeclared here to avoid shadowed, dead definitions. */
   --ease-out-soft: cubic-bezier(0.22, 1, 0.36, 1);
   --ease-spring: cubic-bezier(0.34, 1.3, 0.64, 1);
 
@@ -178,11 +178,9 @@
   --space-8: 2rem;
   --space-12: 3rem;
   --space-16: 4rem;
-  --radius-xs: 0.25rem;
-  --radius-sm: 0.375rem;
-  --radius-md: 0.5rem;
-  --radius-lg: 0.75rem;
-  --radius-xl: 1rem;
+  /* Radius tokens are the single source of truth in @theme above (they also
+     generate the rounded-* utilities); do not redefine them here or var() and
+     the utilities drift apart. */
   --duration-fast: 120ms;
   --duration-base: 180ms;
   --duration-slow: 240ms;

src/components/DashboardFloatingFab.tsx

@@ -1,7 +1,7 @@
 "use client";
 
 import Link from "next/link";
-import { useCallback, useRef, useState } from "react";
+import { useCallback, useEffect, useRef, useState } from "react";
 import { ArrowUp, ClipboardCopy, ExternalLink, Plus, Search, X } from "lucide-react";
 import { cn, floatingControl } from "@/components/ui-primitives";
 
@@ -52,6 +52,10 @@ export function DashboardFloatingFab() {
     setOpen(false);
   }, []);
 
+  // Clear any pending copy-notice timeout on unmount so it can't fire after the
+  // component is gone (leaked timer / stray setState).
+  useEffect(() => clearNotice, [clearNotice]);
+
   return (
     <div className="pointer-events-none fixed inset-x-0 bottom-0 z-40 px-4 pb-6">
       <div className="mx-auto flex w-full max-w-7xl justify-end">

src/lib/evidence.ts

@@ -297,14 +297,18 @@ export function selectBestSourceRecommendation(
   const directQuote = quoteCards.find((quote) => quote.chunk_id === best.id);
   const documentQuote = quoteCards.find((quote) => quote.document_id === best.document_id);
   const quote = directQuote?.quote ?? documentQuote?.quote;
-  const snippet = quote ?? normalizeText(best.content).slice(0, 260).trim();
+  // Track truncation by the pre-slice length rather than guessing from the
+  // post-trim snippet length (=== 260 dropped the ellipsis when trim shortened a
+  // truncated slice, and appended a spurious "..." to a complete 260-char quote).
+  const fullContent = normalizeText(best.content).trim();
+  const snippet = quote ?? (fullContent.length > 260 ? `${fullContent.slice(0, 257).trimEnd()}...` : fullContent);
   const citation = citationFromResult(best);
 
   return {
     ...citation,
     source_strength: best.source_strength ?? sourceStrengthForSimilarity(best.similarity),
     score: best.hybrid_score ?? best.similarity,
-    snippet: snippet.length === 260 ? `${snippet.slice(0, 257).trim()}...` : snippet,
+    snippet,
     quote,
     section_heading: best.section_heading,
     image_count: (best.images ?? []).filter((image) => isClinicalImageEvidence(image)).length,

src/lib/extractors/document.ts

@@ -146,7 +146,24 @@ async function extractDocx(buffer: Buffer) {
       if (file.dir) continue;
       const bytes = await file.async("nodebuffer");
       const ext = path.extname(name).toLowerCase() || ".png";
-      const mimeType = ext === ".jpg" || ext === ".jpeg" ? "image/jpeg" : ext === ".webp" ? "image/webp" : "image/png";
+      // Map the actual media extension to its real MIME type. Previously every
+      // non-jpg/webp extension (including .emf/.wmf/.tiff/.bmp/.gif vector and
+      // legacy-raster figures common in clinical .docx) was mislabeled image/png,
+      // which was then written to storage/DB and sent to the vision API as PNG.
+      const docxImageMimeByExt: Record<string, string> = {
+        ".jpg": "image/jpeg",
+        ".jpeg": "image/jpeg",
+        ".png": "image/png",
+        ".gif": "image/gif",
+        ".webp": "image/webp",
+        ".bmp": "image/bmp",
+        ".tif": "image/tiff",
+        ".tiff": "image/tiff",
+        ".emf": "image/emf",
+        ".wmf": "image/wmf",
+        ".svg": "image/svg+xml",
+      };
+      const mimeType = docxImageMimeByExt[ext] ?? "application/octet-stream";
       const outputPath = path.join(tempRoot, `docx-image-${index}${ext}`);
       await writeFile(outputPath, bytes);
       images.push({

src/lib/rag-answer-text.ts

@@ -8,6 +8,12 @@ const likelyFragmentPhrases =
   /\b(?:answer|heading|body|grounded|confidence|citations?|answerSections?|citation_chunk_ids|conflictsOrGaps|quoteCards?|source_chunk_ids|chunk_id)\b/i;
 const answerSectionArtifactPattern =
   /"?(answer|heading|body|grounded|confidence|citations?|answerSections?|citation_chunk_ids|conflictsOrGaps|quoteCards?|source_chunk_ids|chunk_id)"?\s*:\s*/i;
+// Mid-text truncation must only fire on a genuine leaked JSON key — i.e. a
+// double-quoted key followed by a colon ("confidence":). A bare English word +
+// colon ("...document the confidence: high...") is legitimate clinical prose and
+// must NOT cause the rest of the answer/section to be sliced away.
+const leakedJsonKeyPattern =
+  /"(answer|heading|body|grounded|confidence|citations?|answerSections?|citation_chunk_ids|conflictsOrGaps|quoteCards?|source_chunk_ids|chunk_id)"\s*:/i;
 
 export function normalizeSectionText(value: string) {
   return value.trim().replace(/\s+/g, " ");
@@ -88,11 +94,15 @@ export function sanitizeStructuredText(
   const normalized = normalizeSectionText(sourceTextForClinicalProse(value));
   if (!normalized) return "";
 
+  // A leaked key at the very start is stripped (lenient: it is clearly an
+  // artifact prefix). Mid-text, only truncate at a genuine quoted JSON key so we
+  // never cut real prose at an ordinary word like "confidence:" or "body:".
+  const leakedKeyIndex = normalized.search(leakedJsonKeyPattern);
   const trimmed =
     normalized.search(answerSectionArtifactPattern) === 0
       ? normalized.replace(answerSectionArtifactPattern, "").trim()
-      : normalized.search(answerSectionArtifactPattern) > 0
-        ? normalized.slice(0, normalized.search(answerSectionArtifactPattern)).trim()
+      : leakedKeyIndex > 0
+        ? normalized.slice(0, leakedKeyIndex).trim()
         : normalized;
 
   const finalText = keepLeading ? trimmed : trimmed.trim();

src/lib/rag.ts

@@ -3483,8 +3483,12 @@ function buildExtractiveAnswer(args: {
     }
   }
   for (const quote of quoteCards) {
-    if (!citationIds.has(quote.chunk_id))
-      citations.push(resultCitation(args.results.find((result) => result.id === quote.chunk_id)!));
+    if (!citationIds.has(quote.chunk_id)) {
+      // Guard the lookup: a quote card whose chunk_id was filtered out of results
+      // would make find() return undefined and resultCitation(undefined) throw.
+      const source = args.results.find((result) => result.id === quote.chunk_id);
+      if (source) citations.push(resultCitation(source));
+    }
     citationIds.add(quote.chunk_id);
   }
 
@@ -3506,10 +3510,15 @@ function buildExtractiveAnswer(args: {
     sourceCount: args.results.length,
   });
 
+  // When no concise source sentence could be extracted, buildNaturalExtractiveAnswer
+  // returns a sentinel non-answer with no citationChunkIds. Do not present that as a
+  // grounded, confidence-rated answer just because compactCitations seeded a citation.
+  const hasExtractedAnswer = naturalAnswer.citationChunkIds.length > 0;
+
   return {
     answer: naturalAnswer.answer,
-    grounded: citations.length > 0,
-    confidence: deriveConfidence(args.results, citations.length),
+    grounded: hasExtractedAnswer && citations.length > 0,
+    confidence: hasExtractedAnswer ? deriveConfidence(args.results, citations.length) : "low",
     citations: citations.slice(0, 5),
     sources: args.results,
     modelUsed: null,

src/lib/search-scope.ts

@@ -160,6 +160,10 @@ export async function resolveSearchScope(args: {
       .from("documents")
       .select("id,metadata,import_batch_id")
       .eq("status", "indexed")
+      // Deterministic total order over the unique id column. Without it, separate
+      // LIMIT/OFFSET page queries have no stable order, so rows can be silently
+      // skipped or duplicated across pages and the resolved scope is incomplete.
+      .order("id", { ascending: true })
       .range(offset, Math.min(offset + documentScopeQueryPageSize - 1, maxResolvedDocuments - 1));
     if (args.ownerId) documentQuery = documentQuery.eq("owner_id", args.ownerId);
     if (explicitIds.length) documentQuery = documentQuery.in("id", explicitIds);

tests/eval-cases-route.test.ts

@@ -37,7 +37,7 @@ afterEach(() => {
 describe("/api/eval-cases", () => {
   it("captures a good answer as a promoted eval case and filters malformed chunk ids", async () => {
     const { client, insert } = createInsertMock();
-    vi.doMock("@/lib/env", () => ({ isDemoMode: () => false }));
+    vi.doMock("@/lib/env", () => ({ isDemoMode: () => false, env: { RAG_PERSIST_RAW_QUERY_TEXT: false } }));
     vi.doMock("@/lib/supabase/admin", () => ({ createAdminClient: () => client }));
     vi.doMock("@/lib/supabase/auth", () => ({
       AuthenticationError: class AuthenticationError extends Error {},
@@ -63,6 +63,9 @@ describe("/api/eval-cases", () => {
     expect(response.status).toBe(201);
     expect(payload).toMatchObject({
       owner_id: userId,
+      // PHI control (RET-H4): with raw retention OFF the verbatim query is NOT
+      // stored — only the normalized form.
+      query: "what monitoring is needed for clozapine?",
       query_class: "table_threshold",
       miss_reason: "answer_good_eval",
       top_files: ["CG.MHSP.ClozapinePresAdminMonitor.pdf"],
@@ -77,12 +80,16 @@ describe("/api/eval-cases", () => {
       query_class: "table_threshold",
       source_chunk_ids_rejected: 1,
       cited_chunk_ids_rejected: 1,
+      // Privacy metadata folded in; verbatim answer dropped when retention is off.
+      raw_query_retained: false,
+      query_hash: expect.any(String),
+      answer: "",
     });
   });
 
   it("captures a needs-fixing answer without requiring expected UUID fields", async () => {
     const { client, insert } = createInsertMock();
-    vi.doMock("@/lib/env", () => ({ isDemoMode: () => false }));
+    vi.doMock("@/lib/env", () => ({ isDemoMode: () => false, env: { RAG_PERSIST_RAW_QUERY_TEXT: false } }));
     vi.doMock("@/lib/supabase/admin", () => ({ createAdminClient: () => client }));
     vi.doMock("@/lib/supabase/auth", () => ({
       AuthenticationError: class AuthenticationError extends Error {},
@@ -114,7 +121,7 @@ describe("/api/eval-cases", () => {
 
   it("captures category-specific missed-answer feedback for eval promotion", async () => {
     const { client, insert } = createInsertMock();
-    vi.doMock("@/lib/env", () => ({ isDemoMode: () => false }));
+    vi.doMock("@/lib/env", () => ({ isDemoMode: () => false, env: { RAG_PERSIST_RAW_QUERY_TEXT: false } }));
     vi.doMock("@/lib/supabase/admin", () => ({ createAdminClient: () => client }));
     vi.doMock("@/lib/supabase/auth", () => ({
       AuthenticationError: class AuthenticationError extends Error {},

worker/main.ts

@@ -1507,6 +1507,9 @@ async function processJob(job: JobRow) {
     }
   } finally {
     await cleanupExtractedTemporaryPaths(extracted);
+    // The progress-throttle entry is only needed while this job is processing;
+    // drop it so the module-level map does not grow for the worker's lifetime.
+    progressUpdateState.delete(job.id);
   }
 }