improvement(knowledge): count an unproven reach, read memos under the search deadline, and drop the dead search paths (#8101)

Author: waleedlatif1

apps/sim/app/api/knowledge/search/utils.test.ts

@@ -877,14 +877,4 @@ describe('Knowledge Search Utils', () => {
       )
     })
   })
-
-  describe('getDocumentMetadataByIds', () => {
-    it('should handle empty input gracefully', async () => {
-      const { getDocumentMetadataByIds } = await import('@/lib/knowledge/search/queries')
-
-      const result = await getDocumentMetadataByIds([])
-
-      expect(result).toEqual({})
-    })
-  })
 })

apps/sim/app/api/knowledge/utils.ts

@@ -8,6 +8,7 @@ interface KnowledgeBaseData {
   userId: string
   workspaceId?: string | null
   name: string
+  isSearchIndex: boolean
   description?: string | null
   tokenCount: number
   embeddingModel: string
@@ -22,7 +23,13 @@ export interface KnowledgeBaseAccessResult {
   hasAccess: true
   knowledgeBase: Pick<
     KnowledgeBaseData,
-    'id' | 'userId' | 'workspaceId' | 'name' | 'embeddingModel' | 'embeddingDimension'
+    | 'id'
+    | 'userId'
+    | 'workspaceId'
+    | 'name'
+    | 'isSearchIndex'
+    | 'embeddingModel'
+    | 'embeddingDimension'
   >
 }
 
@@ -52,6 +59,7 @@ async function resolveKnowledgeBaseAccess(
       userId: knowledgeBase.userId,
       workspaceId: knowledgeBase.workspaceId,
       name: knowledgeBase.name,
+      isSearchIndex: knowledgeBase.isSearchIndex,
       embeddingModel: knowledgeBase.embeddingModel,
       embeddingDimension: knowledgeBase.embeddingDimension,
     })

apps/sim/app/api/v1/knowledge/search/route.test.ts

@@ -16,7 +16,6 @@ const {
   mockExecuteKnowledgeSearch,
   mockRetrievalStatus,
   mockGenerateSearchEmbedding,
-  mockGetDocumentMetadataByIds,
   mockGetDocumentTagDefinitions,
   mockAuthenticateRequest,
   mockValidateWorkspaceAccess,
@@ -28,7 +27,6 @@ const {
   mockExecuteKnowledgeSearch: vi.fn(),
   mockRetrievalStatus: vi.fn(() => ({ status: 'complete', timedOutLegs: [] })),
   mockGenerateSearchEmbedding: vi.fn(),
-  mockGetDocumentMetadataByIds: vi.fn(),
   mockGetDocumentTagDefinitions: vi.fn(),
   mockAuthenticateRequest: vi.fn(),
   mockValidateWorkspaceAccess: vi.fn(),
@@ -60,9 +58,7 @@ vi.mock('@/lib/knowledge/search/queries', () => ({
   retrieveKnowledgeSearch: async (params: { access: unknown }) => ({
     rows: await mockExecuteKnowledgeSearch(params),
     retrieval: mockRetrievalStatus(),
-    readAccess: params.access,
   }),
-  getDocumentMetadataByIds: mockGetDocumentMetadataByIds,
 }))
 
 vi.mock('@/app/api/knowledge/utils', () => knowledgeApiUtilsMock)
@@ -132,7 +128,6 @@ describe('v1 knowledge search route — per-KB embedding model', () => {
       isBYOK: false,
     })
     mockExecuteKnowledgeSearch.mockResolvedValue([])
-    mockGetDocumentMetadataByIds.mockResolvedValue({})
     mockGetDocumentTagDefinitions.mockResolvedValue([])
     mockResolveBillingAttribution.mockImplementation(
       ({ actorUserId, workspaceId }: { actorUserId: string; workspaceId: string }) =>
@@ -164,7 +159,6 @@ describe('v1 knowledge search route — per-KB embedding model', () => {
     )
     expect(mockExecuteKnowledgeSearch).toHaveBeenCalledOnce()
     expect(response.status).toBe(500)
-    expect(mockGetDocumentMetadataByIds).not.toHaveBeenCalled()
   })
 
   it('retains the reader provider for ranked results and returned document metadata', async () => {
@@ -193,17 +187,11 @@ describe('v1 knowledge search route — per-KB embedding model', () => {
         accessProvider: provider,
       })
     )
-    expect(mockGetDocumentMetadataByIds).toHaveBeenCalledWith([], access)
   })
 
-  it.each([
-    ['query', false],
-    ['query', true],
-    ['filters', false],
-    ['filters', true],
-  ] as const)(
-    'omits newly denied content from %s results and counts when all denied is %s',
-    async (mode, allDenied) => {
+  it.each(['query', 'filters'] as const)(
+    'renders the source card each %s result row carries',
+    async (mode) => {
       const access = { kind: 'user' as const, userId: 'user-1', tokens: ['reader-token'] }
       const provider = {
         get: vi.fn().mockResolvedValue(access),
@@ -219,26 +207,17 @@ describe('v1 knowledge search route — per-KB embedding model', () => {
         { tagSlot: 'tag1', displayName: 'category', fieldType: 'text' },
       ])
       mockExecuteKnowledgeSearch.mockResolvedValue([
-        {
-          documentId: 'revoked-document',
-          knowledgeBaseId: 'kb-1',
-          content: 'revoked page content',
-          tag1: 'revoked tag',
-          chunkIndex: 0,
-          distance: 0.1,
-        },
         {
           documentId: 'allowed-document',
           knowledgeBaseId: 'kb-1',
           content: 'allowed page content',
+          filename: 'Allowed page',
+          sourceUrl: null,
           tag1: 'docs',
           chunkIndex: 0,
           distance: 0.2,
         },
       ])
-      mockGetDocumentMetadataByIds.mockResolvedValue(
-        allDenied ? {} : { 'allowed-document': { filename: 'Allowed page', sourceUrl: null } }
-      )
       const response = await POST(
         createMockRequest('POST', {
           workspaceId: 'ws-1',
@@ -250,24 +229,19 @@ describe('v1 knowledge search route — per-KB embedding model', () => {
       )
       const body = await response.json()
       expect(response.status).toBe(200)
-      expect(mockGetDocumentMetadataByIds).toHaveBeenCalledWith(
-        ['revoked-document', 'allowed-document'],
-        access
-      )
-      expect(body.data.results).toEqual(
-        allDenied
-          ? []
-          : [
-              expect.objectContaining({
-                documentId: 'allowed-document',
-                documentName: 'Allowed page',
-                content: 'allowed page content',
-                metadata: { category: 'docs' },
-              }),
-            ]
+      expect(mockExecuteKnowledgeSearch).toHaveBeenCalledWith(
+        expect.objectContaining({ access, accessProvider: provider })
       )
-      expect(body.data.totalResults).toBe(allDenied ? 0 : 1)
-      expect(JSON.stringify(body)).not.toContain('revoked')
+      expect(body.data.results).toEqual([
+        expect.objectContaining({
+          documentId: 'allowed-document',
+          documentName: 'Allowed page',
+          sourceUrl: null,
+          content: 'allowed page content',
+          metadata: { category: 'docs' },
+        }),
+      ])
+      expect(body.data.totalResults).toBe(1)
     }
   )
 
@@ -372,7 +346,7 @@ describe('v1 knowledge search route — per-KB embedding model', () => {
     expect(mockGenerateSearchEmbedding).not.toHaveBeenCalled()
   })
 
-  it('surfaces sourceUrl from document metadata in search results', async () => {
+  it('surfaces the sourceUrl a result row carries', async () => {
     mockCheckKnowledgeBaseAccess.mockResolvedValueOnce({
       hasAccess: true,
       knowledgeBase: baseKb('kb-confluence', 'text-embedding-3-small'),
@@ -382,16 +356,12 @@ describe('v1 knowledge search route — per-KB embedding model', () => {
         documentId: 'doc-confluence',
         knowledgeBaseId: 'kb-confluence',
         content: 'page content',
+        filename: 'Runbook.md',
+        sourceUrl: 'https://example.atlassian.net/wiki/spaces/DOCS/pages/12345',
         chunkIndex: 0,
         distance: 0.1,
       },
     ])
-    mockGetDocumentMetadataByIds.mockResolvedValue({
-      'doc-confluence': {
-        filename: 'Runbook.md',
-        sourceUrl: 'https://example.atlassian.net/wiki/spaces/DOCS/pages/12345',
-      },
-    })
 
     const req = createMockRequest('POST', {
       workspaceId: 'ws-1',

apps/sim/app/api/v1/knowledge/search/route.ts

@@ -17,7 +17,6 @@ import {
 import { SearchDeadlineError } from '@/lib/knowledge/search/budget'
 import { resolveKnowledgeSearchDefaults } from '@/lib/knowledge/search/defaults'
 import {
-  getDocumentMetadataByIds,
   type KnowledgeRetrievalResult,
   retrieveKnowledgeSearch,
   type SearchResult,
@@ -267,6 +266,7 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
         accessProvider,
         searchMode,
         boostRecency,
+        searchIndexOnly: accessibleKbs.every((kb) => kb.isSearchIndex),
         query,
         queryVector: {
           vector: JSON.stringify(queryEmbeddingResult.embedding),
@@ -316,14 +316,11 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
     /** v1 cannot express an incomplete search, so a leg that ran out of time fails the request. */
     if (retrieved.retrieval.status === 'partial') throw new SearchDeadlineError()
     const results = retrieved.rows
-    const documentIds = results.map((r) => r.documentId)
-    const documentMetadataMap = await getDocumentMetadataByIds(documentIds, retrieved.readAccess)
-    const readableResults = results.filter((result) => documentMetadataMap[result.documentId])
 
     return NextResponse.json({
       success: true,
       data: {
-        results: readableResults.map((result) => {
+        results: results.map((result) => {
           const kbTagMap = tagDefinitionsMap[result.knowledgeBaseId] || {}
           const tags: Record<string, string | number | boolean | Date | null> = {}
 
@@ -335,11 +332,10 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
             }
           })
 
-          const docMeta = documentMetadataMap[result.documentId]
           return {
             documentId: result.documentId,
-            documentName: docMeta?.filename || undefined,
-            sourceUrl: docMeta?.sourceUrl ?? null,
+            documentName: result.filename || undefined,
+            sourceUrl: result.sourceUrl,
             content: result.content,
             chunkIndex: result.chunkIndex,
             metadata: tags,
@@ -349,7 +345,7 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
         query: query || '',
         knowledgeBaseIds: accessibleKbIds,
         topK,
-        totalResults: readableResults.length,
+        totalResults: results.length,
       },
     })
   } catch (error) {

apps/sim/app/api/v2/knowledge/search/route.provenance.test.ts

@@ -13,7 +13,6 @@ const mocks = vi.hoisted(() => ({
   checkActorUsage: vi.fn(),
   generateEmbedding: vi.fn(),
   executeSearch: vi.fn(),
-  getDocumentMetadata: vi.fn(),
   getTagDefinitions: vi.fn(),
   recordEmbeddingUsage: vi.fn(),
 }))
@@ -66,7 +65,6 @@ vi.mock('@/lib/knowledge/search/queries', () => ({
     rows: await mocks.executeSearch(...args),
     retrieval: { status: 'complete', timedOutLegs: [] },
   }),
-  getDocumentMetadataByIds: mocks.getDocumentMetadata,
 }))
 
 vi.mock('@/lib/knowledge/tags/service', () => ({
@@ -240,9 +238,6 @@ beforeEach(() => {
   mocks.checkActorUsage.mockResolvedValue({ isExceeded: false })
   mocks.generateEmbedding.mockResolvedValue({ embedding: [0.1], isBYOK: false })
   mocks.executeSearch.mockResolvedValue([row])
-  mocks.getDocumentMetadata.mockResolvedValue({
-    'document-1': { filename: 'synthetic.txt', sourceUrl: null },
-  })
   mocks.getTagDefinitions.mockResolvedValue([])
   v2RouteMocks.preauthRate.mockResolvedValue(V2_PREAUTH_RATE_LIMIT_ALLOWED)
   v2RouteMocks.operationRate.mockResolvedValue(V2_OPERATION_RATE_LIMIT_ALLOWED)