@customize-agent/knowledge 4.1.4 → 4.1.5

package/dist/core/index-state-store.d.ts

@@ -140,6 +140,11 @@ export declare class IndexStateStore {
         collectionName?: string;
         relativePath?: string;
     }): number;
+    listChunksSampled(options?: {
+        collectionName?: string;
+        relativePath?: string;
+        sampleSize?: number;
+    }): StoredChunk[];
     listChunksByContentBudget(options?: {
         collectionName?: string;
         relativePath?: string;

package/dist/core/index-state-store.js

@@ -216,6 +216,23 @@ export class IndexStateStore {
         const row = this.db.prepare(`SELECT COUNT(*) AS count FROM kb_chunks ${where}`).get(...params);
         return Number(row?.count || 0);
     }
+    // 均匀步长取样：对超长文件按 chunk_index 取模抽样并强制含最后一块，
+    // 避免 listChunks 前缀取样导致的文件中部/尾部关键信息（工期、质量标准等）覆盖缺失
+    listChunksSampled(options = {}) {
+        const sampleSize = Number(options.sampleSize) || 32;
+        const total = this.countChunks(options);
+        if (total <= sampleSize)
+            return this.listChunks(options);
+        const step = Math.ceil(total / sampleSize);
+        const { where, params } = this.chunkQueryParts(options);
+        const whereClause = where ? `${where} AND` : 'WHERE';
+        const rows = this.db.prepare(`
+      SELECT rowid, * FROM kb_chunks
+      ${whereClause} (chunk_index % ? = 0 OR chunk_index = ?)
+      ORDER BY relative_path, chunk_index
+    `).all(...params, step, total - 1);
+        return rows.map(row => this.rowToChunk(row, 0));
+    }
     listChunksByContentBudget(options = {}) {
         const maxContentChars = Number(options.maxContentChars);
         if (!Number.isFinite(maxContentChars) || maxContentChars <= 0)

package/dist/core/knowledge-base-manager.d.ts

@@ -92,6 +92,10 @@ export declare class KnowledgeBaseManager {
         relativePath?: string;
         limit?: number;
     }): StoredChunk[];
+    listChunksSampled(options?: {
+        relativePath?: string;
+        sampleSize?: number;
+    }): StoredChunk[];
     getFileDetail(relativePath: string, options?: {
         maxChunkContentChars?: number;
     }): {

package/dist/core/knowledge-base-manager.js

@@ -519,6 +519,9 @@ export class KnowledgeBaseManager {
     listChunks(options = {}) {
         return this.store.listChunks(options);
     }
+    listChunksSampled(options = {}) {
+        return this.store.listChunksSampled(options);
+    }
     getFileDetail(relativePath, options = {}) {
         const normalized = this.normalizeRelativePath(relativePath);
         const file = this.store.listRecords().find(record => record.relativePath === normalized);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@customize-agent/knowledge",
-  "version": "4.1.4",
+  "version": "4.1.5",
   "description": "Local knowledge base infrastructure for customize-agent",
   "type": "module",
   "main": "./dist/index.js",