standardized struct and interface impl

Author: irahardianto

.serena/memories/2026-01-05-bug-fixes-report.md

@@ -1,31 +1,8 @@
-# Core Foundation State (Jan 5, 2026)
-
-## Security & Infrastructure
-- **Container Security:** All services (`backend`, `frontend`, `ingestion-worker`) run as non-root users (`appuser`/`nginx`) to prevent container escape.
-- **Network Security:** Frontend Nginx binds to unprivileged port 8080.
-- **Permissions:** Strict file ownership and read/write permissions for application artifacts and shared volumes.
-
-## Ingestion Pipeline Stability
-- **Architecture:** Robust multi-process worker using `pebble`.
-- **Reliability:**
-  - Hard timeouts (30 mins) with auto-kill for stuck processes.
-  - Resource limits (8 CPUs, 8GB RAM) enforced via Docker.
-  - Thread safety enforced via environment variables (`OMP_NUM_THREADS=2`).
-- **Performance:**
-  - Worker: 8 concurrent workers processing files.
-  - Backend: Parallel chunk embedding (5 concurrent routines).
-- **Capabilities:**
-  - Handles large books (e.g., 50MB, 500+ pages).
-  - Extracts rich metadata (Author, Created Date, Page Count).
-  - Standardized on Docling v2 API.
-
-## Code Quality & Testing
-- **Backend Coverage:** High (>90%) for core adapters (`gemini`, `weaviate`) and repositories (`settings`, `source`).
-- **Observability:** Context-aware JSON logging standardized across Backend and Adapters.
-- **Frontend Quality:** Full store and component testing coverage for `jobs`, `stats`, `sources`, and `settings`.
-- **Standards:** STRICT adherence to `technical-constitution` (Dependency Injection, Interface-based design).
-
-## Known Limitations
-- **Ingestion Worker Environment:** Local testing requires manual `pip install pebble`.
-- **Progress Reporting:** "Pending" state is a black box. No page-level progress. (Proposal created: `docs/2026-01-04-proposal-worker-realtime-progression.md`)
-- **Search API:** Metadata (Author, CreatedAt, PageCount, Type) is now returned in `SearchResult` struct and populated by Weaviate. Frontend integration pending.
+The codebase state for `qurio` is significantly ahead of the documentation/bug reports (specifically `2026-01-05-bug-inconsistencies-3.md`). 
+Verified that:
+1. API Envelopes are implemented.
+2. Background Janitor is implemented.
+3. MCP Context Propagation is correct.
+4. TSConfig redundancy is non-existent.
+5. Source Entity timestamp inconsistency (`updated_at` vs `lastSyncedAt`) is resolved across DB, Backend, and Frontend.
+Future tasks should verify code state before assuming bug reports are accurate.

.serena/memories/2026-01-05-test-fixes-report.md

@@ -1,3 +1,10 @@
 - **Logging:** Implemented `apps/backend/internal/logger` package with `ContextHandler` to automatically propagate `correlation_id` from context to JSON logs.
 - **Worker Logic:** Link discovery logic extracted to pure function `DiscoverLinks` in `apps/backend/internal/worker/link_discovery.go` to separate I/O from business logic.
-- **MCP Errors:** `qurio_search` now returns standard JSON-RPC errors (code -32603) for internal failures instead of embedded text errors.
+- **MCP Errors:** `qurio_search` now returns standard JSON-RPC errors (code -32603) for internal failures instead of embedded text errors.
+- **Ingestion Worker:** Refactored `handle_file_task` to return `list[dict]` matching `handle_web_task`, removing brittle manual list wrapping.
+- **Data Consistency:** 
+    - Standardized `Source` entity across stack.
+    - Backend `Source` struct now includes `UpdatedAt` field.
+    - Repository fetches `updated_at` column.
+    - Frontend `Source` interface uses `updated_at` (removed `lastSyncedAt`).
+- **Metadata Exposure:** Promoted metadata fields (`Author`, `CreatedAt`, `PageCount`, `Language`, `Type`, `SourceID`, `URL`) to top-level fields in `SearchResult` struct and refactored Weaviate adapter/MCP handler to use strong typing.

.serena/memories/core_foundation_state.md

@@ -20,7 +20,11 @@ Qurio exposes its knowledge base via the Model Context Protocol (MCP).
 - `filters` (object): Metadata filters (e.g. `type='code'`, `language='go'`).
 
 **Output:**
-- List of results with Title, Type, Language, SourceID, and Content.
+- List of results with strong-typed top-level metadata:
+  - `Content`: The text snippet.
+  - `Title`, `Author`, `CreatedAt`, `PageCount`: Document metadata.
+  - `Type`, `Language`: Content classification.
+  - `SourceID`, `URL`: Origin tracking.
 - Includes explicit instruction: "Use qurio_read_page(url=\"...\") to read the full content of any result."
 
 ### `qurio_list_sources`

.serena/memories/implementation_details.md

@@ -38,7 +38,7 @@ func (r *PostgresRepo) UpdateStatus(ctx context.Context, id, status string) erro
 }
 
 func (r *PostgresRepo) List(ctx context.Context) ([]Source, error) {
-	query := `SELECT id, type, url, status, max_depth, exclusions, name FROM sources WHERE deleted_at IS NULL ORDER BY created_at DESC`
+	query := `SELECT id, type, url, status, max_depth, exclusions, name, updated_at FROM sources WHERE deleted_at IS NULL ORDER BY created_at DESC`
 	rows, err := r.db.QueryContext(ctx, query)
 	if err != nil {
 		return nil, err
@@ -48,7 +48,7 @@ func (r *PostgresRepo) List(ctx context.Context) ([]Source, error) {
 	var sources []Source
 	for rows.Next() {
 		var s Source
-		if err := rows.Scan(&s.ID, &s.Type, &s.URL, &s.Status, &s.MaxDepth, pq.Array(&s.Exclusions), &s.Name); err != nil {
+		if err := rows.Scan(&s.ID, &s.Type, &s.URL, &s.Status, &s.MaxDepth, pq.Array(&s.Exclusions), &s.Name, &s.UpdatedAt); err != nil {
 			return nil, err
 		}
 		sources = append(sources, s)
@@ -58,8 +58,8 @@ func (r *PostgresRepo) List(ctx context.Context) ([]Source, error) {
 
 func (r *PostgresRepo) Get(ctx context.Context, id string) (*Source, error) {
 	s := &Source{}
-	query := `SELECT id, type, url, status, max_depth, exclusions, name FROM sources WHERE id = $1 AND deleted_at IS NULL`
-	err := r.db.QueryRowContext(ctx, query, id).Scan(&s.ID, &s.Type, &s.URL, &s.Status, &s.MaxDepth, pq.Array(&s.Exclusions), &s.Name)
+	query := `SELECT id, type, url, status, max_depth, exclusions, name, updated_at FROM sources WHERE id = $1 AND deleted_at IS NULL`
+	err := r.db.QueryRowContext(ctx, query, id).Scan(&s.ID, &s.Type, &s.URL, &s.Status, &s.MaxDepth, pq.Array(&s.Exclusions), &s.Name, &s.UpdatedAt)
 	if err != nil {
 		return nil, err
 	}

.serena/memories/mcp_tools_spec.md

@@ -44,15 +44,16 @@ func TestPostgresRepo_Get(t *testing.T) {
 
 	repo := source.NewPostgresRepo(db)
 
-	rows := sqlmock.NewRows([]string{"id", "type", "url", "status", "max_depth", "exclusions", "name"}).
-		AddRow("id1", "web", "http://example.com", "active", 2, "{}", "Test")
+	rows := sqlmock.NewRows([]string{"id", "type", "url", "status", "max_depth", "exclusions", "name", "updated_at"}).
+		AddRow("id1", "web", "http://example.com", "active", 2, "{}", "Test", "2024-01-01T00:00:00Z")
 
-	mock.ExpectQuery(regexp.QuoteMeta(`SELECT id, type, url, status, max_depth, exclusions, name FROM sources WHERE id = $1 AND deleted_at IS NULL`)).
+	mock.ExpectQuery(regexp.QuoteMeta(`SELECT id, type, url, status, max_depth, exclusions, name, updated_at FROM sources WHERE id = $1 AND deleted_at IS NULL`)).
 		WithArgs("id1").
 		WillReturnRows(rows)
 
 	s, err := repo.Get(context.Background(), "id1")
 	assert.NoError(t, err)
 	assert.Equal(t, "id1", s.ID)
 	assert.Equal(t, "Test", s.Name)
+	assert.Equal(t, "2024-01-01T00:00:00Z", s.UpdatedAt)
 }

apps/backend/features/source/repo.go

@@ -23,6 +23,7 @@ type Source struct {
 	MaxDepth    int      `json:"max_depth"`
 	Exclusions  []string `json:"exclusions"`
 	Name        string   `json:"name"`
+	UpdatedAt   string   `json:"updated_at"`
 }
 
 type SourcePage struct {

apps/backend/features/source/repo_test.go

@@ -93,4 +93,19 @@ describe('Source Store', () => {
     expect(store.isLoading).toBe(false)
     expect(store.error).toContain('Failed to add source')
   })
+
+  it('should map updated_at correctly', async () => {
+    const store = useSourceStore()
+    // @ts-ignore - simulating API response with field that doesn't exist on type yet
+    const mockSources = [{ id: '1', name: 'Test', updated_at: '2024-01-01' }]
+    
+    global.fetch = vi.fn().mockResolvedValue({
+      ok: true,
+      json: () => Promise.resolve({ data: mockSources })
+    })
+
+    await store.fetchSources()
+    // This assertion relies on the field existing on the interface
+    expect(store.sources[0].updated_at).toBe('2024-01-01')
+  })
 })

apps/backend/features/source/source.go

@@ -19,7 +19,7 @@ export interface Source {
   type?: string
   url?: string
   status?: string
-  lastSyncedAt?: string
+  updated_at?: string
   max_depth?: number
   exclusions?: string[]
   chunks?: Chunk[]