fix(server): frequent embedding (#13475)

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

- New Features
- Smarter embedding pipeline skips re-embedding when content hasn’t
changed; added content sanitization for embeddings and workspace content
retrieval.
- Bug Fixes
- Re-embedding now requires both a document update and the last
embedding being older than 10 minutes, reducing unnecessary work.
- Refactor
- Consolidated embedding preprocessing and moved sanitization utilities
into shared models; upserts now refresh stored content.
- Tests
- Expanded snapshot-based tests covering multiple time/age scenarios for
embedding decision logic.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

packages/backend/server/src/models/common/copilot.ts

@@ -148,3 +148,36 @@ export type IgnoredDoc = {
   createdByAvatar: string | undefined;
   updatedBy: string | undefined;
 };
+
+export const EMBEDDING_DIMENSIONS = 1024;
+
+const FILTER_PREFIX = [
+  'Title: ',
+  'Created at: ',
+  'Updated at: ',
+  'Created by: ',
+  'Updated by: ',
+];
+
+export function clearEmbeddingContent(content: string): string {
+  const lines = content.split('\n');
+  let maxLines = 5;
+  while (maxLines > 0 && lines.length > 0) {
+    if (FILTER_PREFIX.some(prefix => lines[0].startsWith(prefix))) {
+      lines.shift();
+      maxLines--;
+    } else {
+      // only process consecutive metadata rows
+      break;
+    }
+  }
+  return lines.join('\n');
+}
+
+export function clearEmbeddingChunk(chunk: ChunkSimilarity): ChunkSimilarity {
+  if (chunk.content) {
+    const content = clearEmbeddingContent(chunk.content);
+    return { ...chunk, content };
+  }
+  return chunk;
+}

packages/backend/server/src/models/copilot-context.ts

@@ -6,6 +6,7 @@ import { Prisma } from '@prisma/client';
 import { CopilotSessionNotFound } from '../base';
 import { BaseModel } from './base';
 import {
+  clearEmbeddingContent,
   ContextBlob,
   ContextConfigSchema,
   ContextDoc,
@@ -13,14 +14,13 @@ import {
   CopilotContext,
   DocChunkSimilarity,
   Embedding,
+  EMBEDDING_DIMENSIONS,
   FileChunkSimilarity,
   MinimalContextConfigSchema,
 } from './common/copilot';
 
 type UpdateCopilotContextInput = Pick<CopilotContext, 'config'>;
 
-export const EMBEDDING_DIMENSIONS = 1024;
-
 /**
  * Copilot Job Model
  */
@@ -215,8 +215,9 @@ export class CopilotContextModel extends BaseModel {
       select: { content: true },
       orderBy: { chunk: 'asc' },
     });
-    return file?.map(f => f.content).join('\n');
+    return file?.map(f => clearEmbeddingContent(f.content)).join('\n');
   }
+
   async insertFileEmbedding(
     contextId: string,
     fileId: string,
@@ -263,6 +264,19 @@ export class CopilotContextModel extends BaseModel {
     return similarityChunks.filter(c => Number(c.distance) <= threshold);
   }
 
+  async getWorkspaceContent(
+    workspaceId: string,
+    docId: string,
+    chunk?: number
+  ): Promise<string | undefined> {
+    const file = await this.db.aiWorkspaceEmbedding.findMany({
+      where: { workspaceId, docId, chunk },
+      select: { content: true },
+      orderBy: { chunk: 'asc' },
+    });
+    return file?.map(f => clearEmbeddingContent(f.content)).join('\n');
+  }
+
   async insertWorkspaceEmbedding(
     workspaceId: string,
     docId: string,
@@ -287,6 +301,7 @@ export class CopilotContextModel extends BaseModel {
       VALUES ${values}
       ON CONFLICT (workspace_id, doc_id, chunk)
       DO UPDATE SET
+        content = EXCLUDED.content,
         embedding = EXCLUDED.embedding,
         updated_at = excluded.updated_at;
     `;

packages/backend/server/src/models/copilot-workspace.ts

@@ -242,10 +242,9 @@ export class CopilotWorkspaceConfigModel extends BaseModel {
   @Transactional()
   async checkDocNeedEmbedded(workspaceId: string, docId: string) {
     // NOTE: check if the document needs re-embedding.
-    // 1. check if there have been any recent updates to the document snapshot and update
-    // 2. check if the embedding is older than the snapshot and update
-    // 3. check if the embedding is older than 10 minutes (avoid frequent updates)
-    // if all conditions are met, re-embedding is required.
+    // 1. first-time embedding when no embedding exists
+    // 2. re-embedding only when the doc has updates newer than the last embedding
+    //    AND the last embedding is older than 10 minutes (avoid frequent updates)
     const result = await this.db.$queryRaw<{ needs_embedding: boolean }[]>`
       SELECT
         EXISTS (
@@ -280,8 +279,7 @@ export class CopilotWorkspaceConfigModel extends BaseModel {
               AND e.doc_id = docs.doc_id
           WHERE
             e.updated_at IS NULL
-            OR docs.updated_at > e.updated_at
-            OR e.updated_at < NOW() - INTERVAL '10 minutes'
+            OR (docs.updated_at > e.updated_at AND e.updated_at < NOW() - INTERVAL '10 minutes')
         ) AS needs_embedding;
     `;