@gmickel/gno 1.8.0 → 1.9.0

package/src/pipeline/hybrid.ts

@@ -841,6 +841,8 @@ export async function searchHybrid(
       score: candidate.blendedScore,
       uri: doc.uri,
       title: doc.title ?? undefined,
+      contentType: doc.contentType ?? undefined,
+      categories: doc.categories ?? undefined,
       line: snippetChunk.startLine,
       snippet,
       snippetLanguage: chunk.language ?? undefined,

package/src/pipeline/search.ts

@@ -117,6 +117,8 @@ function buildSearchResult(ctx: BuildResultContext): SearchResult {
     score: fts.score, // Raw score, normalized later as batch
     uri: fts.uri ?? "",
     title: fts.title,
+    contentType: fts.contentType,
+    categories: fts.categories,
     line: chunk?.startLine,
     snippet,
     snippetLanguage: chunk?.language ?? undefined,

package/src/pipeline/types.ts

@@ -43,6 +43,8 @@ export interface SearchResult {
   score: number;
   uri: string;
   title?: string;
+  contentType?: string;
+  categories?: string[];
   /** Best source line for editor/agent anchors (1-indexed) */
   line?: number;
   snippet: string;

package/src/pipeline/vsearch.ts

@@ -227,6 +227,8 @@ export async function searchVectorWithEmbedding(
       score,
       uri: doc.uri,
       title: doc.title ?? undefined,
+      contentType: doc.contentType ?? undefined,
+      categories: doc.categories ?? undefined,
       line: chunk.startLine,
       snippet: chunk.text,
       snippetLanguage: chunk.language ?? undefined,
@@ -281,6 +283,8 @@ export async function searchVectorWithEmbedding(
         score,
         uri: doc.uri,
         title: doc.title ?? undefined,
+        contentType: doc.contentType ?? undefined,
+        categories: doc.categories ?? undefined,
         line: chunk.startLine,
         snippet: fullContent ?? chunk.text,
         snippetLanguage: chunk.language ?? undefined,

package/src/sdk/client.ts

@@ -44,7 +44,11 @@ import {
   getIndexDbPath,
   parseUri,
 } from "../app/constants";
-import { ConfigSchema, loadConfig } from "../config";
+import {
+  ConfigSchema,
+  loadConfig,
+  normalizeConfigContentTypes,
+} from "../config";
 import {
   buildCaptureReceipt,
   type CapturePlan,
@@ -70,7 +74,11 @@ import { resolveNotePreset } from "../core/note-presets";
 import { extractSections } from "../core/sections";
 import { normalizeStructuredQueryInput } from "../core/structured-query";
 import { parseAndValidateTagFilter } from "../core/tags";
-import { defaultSyncService, type SyncResult } from "../ingestion";
+import {
+  defaultSyncService,
+  type SyncResult,
+  withContentTypeRules,
+} from "../ingestion";
 import { updateFrontmatterTags } from "../ingestion/frontmatter";
 import { LlmAdapter } from "../llm/nodeLlamaCpp/adapter";
 import { resolveDownloadPolicy } from "../llm/policy";
@@ -141,7 +149,7 @@ async function resolveClientState(
         parsed.error.issues[0]?.message ?? "Invalid config"
       );
     }
-    config = parsed.data;
+    config = normalizeConfigContentTypes(parsed.data).config;
     configPath = null;
     configSource = "inline";
   } else {
@@ -695,10 +703,17 @@ class GnoClientImpl implements GnoClient {
   async update(options: GnoUpdateOptions = {}): Promise<SyncResult> {
     this.assertOpen();
     const collections = this.getCollections(options.collection);
-    return defaultSyncService.syncAll(collections, this.store, {
-      gitPull: options.gitPull,
-      runUpdateCmd: true,
-    });
+    return defaultSyncService.syncAll(
+      collections,
+      this.store,
+      withContentTypeRules(
+        {
+          gitPull: options.gitPull,
+          runUpdateCmd: true,
+        },
+        this.config
+      )
+    );
   }
 
   async embed(options: GnoEmbedOptions = {}): Promise<GnoEmbedResult> {
@@ -808,10 +823,13 @@ class GnoClientImpl implements GnoClient {
       collection,
       this.store,
       [plan.relPath],
-      {
-        runUpdateCmd: false,
-        gitPull: false,
-      }
+      withContentTypeRules(
+        {
+          runUpdateCmd: false,
+          gitPull: false,
+        },
+        this.config
+      )
     );
     const syncResult = syncResults[0];
     if (!syncResult || syncResult.status === "error") {
@@ -903,10 +921,13 @@ class GnoClientImpl implements GnoClient {
       collection,
       this.store,
       [plan.relPath],
-      {
-        runUpdateCmd: false,
-        gitPull: false,
-      }
+      withContentTypeRules(
+        {
+          runUpdateCmd: false,
+          gitPull: false,
+        },
+        this.config
+      )
     );
     const syncResult = syncResults[0];
     const docResult = await this.store.getDocument(
@@ -987,9 +1008,11 @@ class GnoClientImpl implements GnoClient {
     const currentPath = `${collection.path}/${storedDoc.relPath}`;
     const nextPath = `${collection.path}/${plan.nextRelPath}`;
     await renameFilePath(currentPath, nextPath);
-    await defaultSyncService.syncCollection(collection, this.store, {
-      runUpdateCmd: false,
-    });
+    await defaultSyncService.syncCollection(
+      collection,
+      this.store,
+      withContentTypeRules({ runUpdateCmd: false }, this.config)
+    );
     const linksResult = await this.store.getLinksForDoc(storedDoc.id);
     const backlinksResult = await this.store.getBacklinksForDoc(storedDoc.id);
     if (!linksResult.ok || !backlinksResult.ok) {
@@ -1044,9 +1067,11 @@ class GnoClientImpl implements GnoClient {
     const nextPath = `${collection.path}/${plan.nextRelPath}`;
     await mkdir(dirname(nextPath), { recursive: true });
     await renameFilePath(currentPath, nextPath);
-    await defaultSyncService.syncCollection(collection, this.store, {
-      runUpdateCmd: false,
-    });
+    await defaultSyncService.syncCollection(
+      collection,
+      this.store,
+      withContentTypeRules({ runUpdateCmd: false }, this.config)
+    );
     const linksResult = await this.store.getLinksForDoc(storedDoc.id);
     const backlinksResult = await this.store.getBacklinksForDoc(storedDoc.id);
     if (!linksResult.ok || !backlinksResult.ok) {
@@ -1113,9 +1138,11 @@ class GnoClientImpl implements GnoClient {
     const nextPath = `${collection.path}/${plan.nextRelPath}`;
     await mkdir(dirname(nextPath), { recursive: true });
     await copyFilePath(currentPath, nextPath);
-    await defaultSyncService.syncCollection(collection, this.store, {
-      runUpdateCmd: false,
-    });
+    await defaultSyncService.syncCollection(
+      collection,
+      this.store,
+      withContentTypeRules({ runUpdateCmd: false }, this.config)
+    );
     const linksResult = await this.store.getLinksForDoc(storedDoc.id);
     const backlinksResult = await this.store.getBacklinksForDoc(storedDoc.id);
     if (!linksResult.ok || !backlinksResult.ok) {

package/src/serve/background-runtime.ts

@@ -11,11 +11,13 @@ import type {
 import { getIndexDbPath } from "../app/constants";
 import {
   ensureDirectories,
+  formatConfigWarnings,
   getConfigPaths,
   isInitialized,
   loadConfig,
 } from "../config";
 import { defaultSyncService } from "../ingestion";
+import { withContentTypeRules } from "../ingestion";
 import { getActivePreset } from "../llm/registry";
 import { SqliteAdapter } from "../store/sqlite/adapter";
 import {
@@ -79,6 +81,7 @@ type BackgroundRuntimeDeps = {
     scheduler: EmbedScheduler | null;
     eventBus?: DocumentEventBus | null;
     callbacks?: CollectionWatchCallbacks;
+    syncOptions?: Parameters<typeof withContentTypeRules>[0];
   }) => CollectionWatchService;
 };
 
@@ -103,6 +106,9 @@ export async function startBackgroundRuntime(
   if (!configResult.ok) {
     return { success: false, error: configResult.error.message };
   }
+  for (const warning of formatConfigWarnings(configResult.warnings)) {
+    console.warn(warning);
+  }
   const config = configResult.value;
 
   if (options.requireCollections && config.collections.length === 0) {
@@ -171,6 +177,7 @@ export async function startBackgroundRuntime(
     scheduler,
     eventBus: options.eventBus ?? null,
     callbacks: options.watchCallbacks,
+    syncOptions: withContentTypeRules({}, config),
   });
   watchService.start();
   ctxHolder.watchService = watchService;
@@ -189,10 +196,17 @@ export async function startBackgroundRuntime(
       eventBus: options.eventBus ?? null,
       watchService,
       async syncAll(syncOptions = {}) {
-        const syncResult = await syncAllService(config.collections, store, {
-          gitPull: syncOptions.gitPull,
-          runUpdateCmd: syncOptions.runUpdateCmd,
-        });
+        const syncResult = await syncAllService(
+          config.collections,
+          store,
+          withContentTypeRules(
+            {
+              gitPull: syncOptions.gitPull,
+              runUpdateCmd: syncOptions.runUpdateCmd,
+            },
+            config
+          )
+        );
 
         let embedResult: EmbedResult | null = null;
         if (syncOptions.triggerEmbed !== false) {

package/src/serve/config-sync.ts

@@ -14,6 +14,7 @@ import {
   type ApplyConfigResult,
   type MutationResult,
 } from "../core/config-mutation";
+import { withContentTypeRules } from "../ingestion";
 
 /**
  * Apply a config mutation atomically with serialization.
@@ -53,7 +54,10 @@ export async function applyConfigChange(
   );
 
   if (result.ok) {
-    ctxHolder.watchService?.updateCollections(ctxHolder.config.collections);
+    ctxHolder.watchService?.updateCollections(
+      ctxHolder.config.collections,
+      withContentTypeRules({}, ctxHolder.config)
+    );
   }
 
   return result;
@@ -78,7 +82,10 @@ export async function applyConfigChangeTyped<T>(
   );
 
   if (result.ok) {
-    ctxHolder.watchService?.updateCollections(ctxHolder.config.collections);
+    ctxHolder.watchService?.updateCollections(
+      ctxHolder.config.collections,
+      withContentTypeRules({}, ctxHolder.config)
+    );
   }
 
   return result;

package/src/serve/routes/api.ts

@@ -82,7 +82,11 @@ import {
   validateTag,
 } from "../../core/tags";
 import { validateRelPath } from "../../core/validation";
-import { defaultSyncService, type SyncResult } from "../../ingestion";
+import {
+  defaultSyncService,
+  type SyncResult,
+  withContentTypeRules,
+} from "../../ingestion";
 import { updateFrontmatterTags } from "../../ingestion/frontmatter";
 import {
   getCollectionEffectiveModels,
@@ -919,10 +923,17 @@ export async function handleCreateCollection(
     return errorResponse("RUNTIME", "Collection not found after add", 500);
   }
   const jobResult = startJob("add", async (): Promise<SyncResult> => {
-    const result = await defaultSyncService.syncCollection(collection, store, {
-      gitPull: body.gitPull,
-      runUpdateCmd: true,
-    });
+    const result = await defaultSyncService.syncCollection(
+      collection,
+      store,
+      withContentTypeRules(
+        {
+          gitPull: body.gitPull,
+          runUpdateCmd: true,
+        },
+        syncResult.config
+      )
+    );
     if (result.filesAdded > 0 || result.filesUpdated > 0) {
       if (ctxHolder.scheduler) {
         await ctxHolder.scheduler.triggerNow();
@@ -1214,10 +1225,17 @@ export async function handleSync(
 
   // Start background sync job
   const jobResult = startJob("sync", async (): Promise<SyncResult> => {
-    const result = await defaultSyncService.syncAll(collections, store, {
-      gitPull: body.gitPull,
-      runUpdateCmd: true,
-    });
+    const result = await defaultSyncService.syncAll(
+      collections,
+      store,
+      withContentTypeRules(
+        {
+          gitPull: body.gitPull,
+          runUpdateCmd: true,
+        },
+        ctxHolder.config
+      )
+    );
     if (result.totalFilesAdded > 0 || result.totalFilesUpdated > 0) {
       if (ctxHolder.scheduler) {
         await ctxHolder.scheduler.triggerNow();
@@ -1843,9 +1861,11 @@ export async function handleRenameDoc(
     const nextUri = `gno://${collection.name}/${nextRelPath}`;
     let warning: string | undefined;
     try {
-      await syncCollection(collection, store, {
-        runUpdateCmd: false,
-      });
+      await syncCollection(
+        collection,
+        store,
+        withContentTypeRules({ runUpdateCmd: false }, ctxHolder.config)
+      );
     } catch {
       warning =
         "File renamed on disk, but index refresh failed. Run Update All to reconcile the workspace.";
@@ -2063,9 +2083,11 @@ export async function handleTrashDoc(
     }
     let warning: string | undefined;
     try {
-      await syncCollection(collection, store, {
-        runUpdateCmd: false,
-      });
+      await syncCollection(
+        collection,
+        store,
+        withContentTypeRules({ runUpdateCmd: false }, ctxHolder.config)
+      );
     } catch {
       warning =
         "File moved to Trash, but index refresh failed. Run Update All to reconcile the workspace.";
@@ -2184,9 +2206,11 @@ export async function handleMoveDoc(
     await renameFilePath(fullPath, nextFullPath);
     let warning: string | undefined;
     try {
-      await defaultSyncService.syncCollection(collection, store, {
-        runUpdateCmd: false,
-      });
+      await defaultSyncService.syncCollection(
+        collection,
+        store,
+        withContentTypeRules({ runUpdateCmd: false }, ctxHolder.config)
+      );
     } catch {
       warning =
         "File moved on disk, but index refresh failed. Run Update All to reconcile the workspace.";
@@ -2305,9 +2329,11 @@ export async function handleDuplicateDoc(
     await copyFilePath(fullPath, nextFullPath);
     let warning: string | undefined;
     try {
-      await defaultSyncService.syncCollection(collection, store, {
-        runUpdateCmd: false,
-      });
+      await defaultSyncService.syncCollection(
+        collection,
+        store,
+        withContentTypeRules({ runUpdateCmd: false }, ctxHolder.config)
+      );
     } catch {
       warning =
         "File duplicated on disk, but index refresh failed. Run Update All to reconcile the workspace.";
@@ -2636,7 +2662,7 @@ export async function handleUpdateDoc(
         const result = await defaultSyncService.syncCollection(
           collection,
           store,
-          { runUpdateCmd: false }
+          withContentTypeRules({ runUpdateCmd: false }, ctxHolder.config)
         );
         // Notify scheduler after sync completes
         ctxHolder.scheduler?.notifySyncComplete([doc.docid]);
@@ -2904,7 +2930,7 @@ export async function handleCreateCapture(
       const result = await defaultSyncService.syncCollection(
         collection,
         store,
-        { runUpdateCmd: false }
+        withContentTypeRules({ runUpdateCmd: false }, ctxHolder.config)
       );
       ctxHolder.scheduler?.notifySyncComplete([plan.relPath]);
       ctxHolder.eventBus?.emit({
@@ -3140,7 +3166,7 @@ export async function handleCreateDoc(
       const result = await defaultSyncService.syncCollection(
         collection,
         store,
-        { runUpdateCmd: false }
+        withContentTypeRules({ runUpdateCmd: false }, ctxHolder.config)
       );
       // Notify scheduler after sync completes (use gnoUri as docid placeholder)
       // The sync will create a proper docid, but we don't have it here yet

package/src/serve/watch-service.ts

@@ -2,7 +2,7 @@ import { watch, type FSWatcher } from "node:fs";
 import { join, normalize, sep } from "node:path";
 
 import type { Collection } from "../config/types";
-import type { CollectionSyncResult } from "../ingestion";
+import type { CollectionSyncResult, SyncOptions } from "../ingestion";
 import type { SqliteAdapter } from "../store/sqlite/adapter";
 import type { DocumentEvent, DocumentEventBus } from "./doc-events";
 import type { EmbedScheduler } from "./embed-scheduler";
@@ -39,6 +39,7 @@ interface CollectionWatchServiceOptions {
   scheduler: EmbedScheduler | null;
   eventBus?: DocumentEventBus | null;
   callbacks?: CollectionWatchCallbacks;
+  syncOptions?: SyncOptions;
   watchFactory?: typeof watch;
 }
 
@@ -48,6 +49,7 @@ export class CollectionWatchService {
   readonly #scheduler: EmbedScheduler | null;
   readonly #eventBus: DocumentEventBus | null;
   readonly #callbacks: CollectionWatchCallbacks | null;
+  #syncOptions: SyncOptions;
   readonly #watchers = new Map<string, FSWatcher>();
   readonly #pendingByCollection = new Map<string, Set<string>>();
   readonly #timers = new Map<string, ReturnType<typeof setTimeout>>();
@@ -64,6 +66,7 @@ export class CollectionWatchService {
     this.#scheduler = options.scheduler;
     this.#eventBus = options.eventBus ?? null;
     this.#callbacks = options.callbacks ?? null;
+    this.#syncOptions = options.syncOptions ?? {};
     this.#watchFactory = options.watchFactory ?? watch;
   }
 
@@ -71,8 +74,14 @@ export class CollectionWatchService {
     this.updateCollections(this.#collections);
   }
 
-  updateCollections(collections: Collection[]): void {
+  updateCollections(
+    collections: Collection[],
+    syncOptions?: SyncOptions
+  ): void {
     this.#collections = collections;
+    if (syncOptions) {
+      this.#syncOptions = syncOptions;
+    }
     const nextNames = new Set(collections.map((collection) => collection.name));
 
     for (const [collectionName, watcher] of this.#watchers) {
@@ -204,6 +213,7 @@ export class CollectionWatchService {
         collection,
         this.#store,
         {
+          ...this.#syncOptions,
           runUpdateCmd: false,
         }
       );

package/src/store/migrations/009-content-type-rule-fingerprint.ts

@@ -0,0 +1,36 @@
+/**
+ * Migration: content type rule fingerprint.
+ *
+ * @module src/store/migrations/009-content-type-rule-fingerprint
+ */
+
+import type { Migration } from "./runner";
+
+export const migration: Migration = {
+  version: 9,
+  name: "content_type_rule_fingerprint",
+
+  up(db): void {
+    const columns = new Set(
+      db
+        .query<{ name: string }, []>("PRAGMA table_info(documents)")
+        .all()
+        .map((row) => row.name)
+    );
+    if (!columns.has("content_type_rules_fingerprint")) {
+      db.exec(
+        "ALTER TABLE documents ADD COLUMN content_type_rules_fingerprint TEXT"
+      );
+    }
+    db.exec(
+      "CREATE INDEX IF NOT EXISTS idx_documents_content_type_rules_fingerprint ON documents(content_type_rules_fingerprint)"
+    );
+  },
+
+  down(db): void {
+    db.exec(
+      "DROP INDEX IF EXISTS idx_documents_content_type_rules_fingerprint"
+    );
+    // SQLite cannot drop columns without a table rebuild; keep column on rollback.
+  },
+};

package/src/store/migrations/index.ts

@@ -22,6 +22,17 @@ import { migration as m005 } from "./005-graph-indexes";
 import { migration as m006 } from "./006-document-metadata";
 import { migration as m007 } from "./007-document-date-fields";
 import { migration as m008 } from "./008-vector-fingerprints";
+import { migration as m009 } from "./009-content-type-rule-fingerprint";
 
 /** All migrations in order */
-export const migrations = [m001, m002, m003, m004, m005, m006, m007, m008];
+export const migrations = [
+  m001,
+  m002,
+  m003,
+  m004,
+  m005,
+  m006,
+  m007,
+  m008,
+  m009,
+];

package/src/store/sqlite/adapter.ts

@@ -554,9 +554,10 @@ export class SqliteAdapter implements StorePort, SqliteDbProvider {
           collection, rel_path, source_hash, source_mime, source_ext,
           source_size, source_mtime, source_ctime, docid, uri, title, mirror_hash,
           converter_id, converter_version, language_hint, content_type, categories,
-          author, frontmatter_date, date_fields, active, indexed_at, last_error_code,
-          last_error_message, last_error_at, ingest_version, updated_at
-        ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, 1, datetime('now'), ?, ?, ?, ?, datetime('now'))
+          author, frontmatter_date, date_fields, content_type_rules_fingerprint,
+          active, indexed_at, last_error_code, last_error_message, last_error_at,
+          ingest_version, updated_at
+        ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, 1, datetime('now'), ?, ?, ?, ?, datetime('now'))
         ON CONFLICT(collection, rel_path) DO UPDATE SET
           source_hash = excluded.source_hash,
           source_mime = excluded.source_mime,
@@ -576,6 +577,7 @@ export class SqliteAdapter implements StorePort, SqliteDbProvider {
           author = excluded.author,
           frontmatter_date = excluded.frontmatter_date,
           date_fields = excluded.date_fields,
+          content_type_rules_fingerprint = excluded.content_type_rules_fingerprint,
           active = 1,
           indexed_at = datetime('now'),
           last_error_code = excluded.last_error_code,
@@ -605,6 +607,7 @@ export class SqliteAdapter implements StorePort, SqliteDbProvider {
           doc.author ?? null,
           doc.frontmatterDate ?? null,
           doc.dateFields ? JSON.stringify(doc.dateFields) : null,
+          doc.contentTypeRulesFingerprint ?? null,
           doc.lastErrorCode ?? null,
           doc.lastErrorMessage ?? null,
           doc.lastErrorCode ? new Date().toISOString() : null,
@@ -1366,7 +1369,9 @@ export class SqliteAdapter implements StorePort, SqliteDbProvider {
           d.source_mtime,
           d.frontmatter_date,
           d.source_size,
-          d.source_hash
+          d.source_hash,
+          d.content_type,
+          d.categories
         FROM fts_matches fm
         JOIN documents d ON d.id = fm.rowid AND d.active = 1
         WHERE 1 = 1
@@ -1392,6 +1397,8 @@ export class SqliteAdapter implements StorePort, SqliteDbProvider {
         frontmatter_date: string | null;
         source_size: number | null;
         source_hash: string | null;
+        content_type: string | null;
+        categories: string | null;
       }
 
       const queryParams = [builtQuery.query, ftsLimit, ...params];
@@ -1416,6 +1423,8 @@ export class SqliteAdapter implements StorePort, SqliteDbProvider {
           frontmatterDate: r.frontmatter_date ?? undefined,
           sourceSize: r.source_size ?? undefined,
           sourceHash: r.source_hash ?? undefined,
+          contentType: r.content_type ?? undefined,
+          categories: parseCategoriesJson(r.categories) ?? undefined,
         }))
       );
     } catch (cause) {
@@ -3522,6 +3531,7 @@ interface DbDocumentRow {
   author: string | null;
   frontmatter_date: string | null;
   date_fields: string | null;
+  content_type_rules_fingerprint: string | null;
   indexed_at: string | null;
   active: number;
   ingest_version: number | null;
@@ -3580,19 +3590,23 @@ function mapContextRow(row: DbContextRow): ContextRow {
   };
 }
 
-function mapDocumentRow(row: DbDocumentRow): DocumentRow {
-  let categories: string[] | null = null;
-  if (row.categories) {
-    try {
-      const parsed = JSON.parse(row.categories);
-      if (Array.isArray(parsed)) {
-        categories = parsed.filter((v): v is string => typeof v === "string");
-      }
-    } catch {
-      categories = null;
+function parseCategoriesJson(raw: string | null): string[] | null {
+  if (!raw) {
+    return null;
+  }
+  try {
+    const parsed = JSON.parse(raw);
+    if (Array.isArray(parsed)) {
+      return parsed.filter((v): v is string => typeof v === "string");
     }
+  } catch {
+    return null;
   }
+  return null;
+}
 
+function mapDocumentRow(row: DbDocumentRow): DocumentRow {
+  const categories = parseCategoriesJson(row.categories);
   let dateFields: Record<string, string> | null = null;
   if (row.date_fields) {
     try {
@@ -3638,6 +3652,7 @@ function mapDocumentRow(row: DbDocumentRow): DocumentRow {
     indexedAt: row.indexed_at,
     active: row.active === 1,
     ingestVersion: row.ingest_version,
+    contentTypeRulesFingerprint: row.content_type_rules_fingerprint,
     lastErrorCode: row.last_error_code,
     lastErrorMessage: row.last_error_message,
     lastErrorAt: row.last_error_at,

package/src/store/types.ts

@@ -118,6 +118,8 @@ export interface DocumentRow {
   active: boolean;
   /** Ingest schema version for backfill detection */
   ingestVersion: number | null;
+  /** Fingerprint of normalized content type rules used for derived metadata */
+  contentTypeRulesFingerprint?: string | null;
 
   // Error tracking
   lastErrorCode: string | null;
@@ -264,6 +266,8 @@ export interface DocumentInput {
   lastErrorMessage?: string;
   /** Ingest schema version for backfill detection */
   ingestVersion?: number;
+  /** Fingerprint of normalized content type rules used for derived metadata */
+  contentTypeRulesFingerprint?: string;
 }
 
 /** Result of upserting a document */
@@ -345,6 +349,8 @@ export interface FtsResult {
   frontmatterDate?: string;
   sourceSize?: number;
   sourceHash?: string;
+  contentType?: string;
+  categories?: string[];
 }
 
 // ─────────────────────────────────────────────────────────────────────────────