From 676d41d4156c5089e84a47da5024cc0071a1a250 Mon Sep 17 00:00:00 2001
From: Peter Steinberger <steipete@gmail.com>
Date: Sun, 18 Jan 2026 09:28:42 +0000
Subject: [PATCH] fix: seed embedding cache for atomic reindex

---
 ...onboard-non-interactive.ai-gateway.test.ts |  7 ++-
 .../onboard-non-interactive.token.test.ts     |  7 ++-
 src/memory/manager.atomic-reindex.test.ts     |  1 +
 src/memory/manager.ts                         | 57 +++++++++++++++++++
 4 files changed, 66 insertions(+), 6 deletions(-)

diff --git a/src/commands/onboard-non-interactive.ai-gateway.test.ts b/src/commands/onboard-non-interactive.ai-gateway.test.ts
index 14afec665..fbd9b08dc 100644
--- a/src/commands/onboard-non-interactive.ai-gateway.test.ts
+++ b/src/commands/onboard-non-interactive.ai-gateway.test.ts
@@ -2,7 +2,7 @@ import fs from "node:fs/promises";
 import os from "node:os";
 import path from "node:path";
 
-import { describe, expect, it } from "vitest";
+import { describe, expect, it, vi } from "vitest";
 
 describe("onboard (non-interactive): Vercel AI Gateway", () => {
   it("stores the API key and configures the default model", async () => {
@@ -27,8 +27,9 @@ describe("onboard (non-interactive): Vercel AI Gateway", () => {
 
     const tempHome = await fs.mkdtemp(path.join(os.tmpdir(), "clawdbot-onboard-gateway-"));
     process.env.HOME = tempHome;
-    delete process.env.CLAWDBOT_STATE_DIR;
-    delete process.env.CLAWDBOT_CONFIG_PATH;
+    process.env.CLAWDBOT_STATE_DIR = tempHome;
+    process.env.CLAWDBOT_CONFIG_PATH = path.join(tempHome, "clawdbot.json");
+    vi.resetModules();
 
     const runtime = {
       log: () => {},
diff --git a/src/commands/onboard-non-interactive.token.test.ts b/src/commands/onboard-non-interactive.token.test.ts
index 28b5a3aa3..f2455ff24 100644
--- a/src/commands/onboard-non-interactive.token.test.ts
+++ b/src/commands/onboard-non-interactive.token.test.ts
@@ -2,7 +2,7 @@ import fs from "node:fs/promises";
 import os from "node:os";
 import path from "node:path";
 
-import { describe, expect, it } from "vitest";
+import { describe, expect, it, vi } from "vitest";
 
 describe("onboard (non-interactive): token auth", () => {
   it("writes token profile config and stores the token", async () => {
@@ -27,8 +27,9 @@ describe("onboard (non-interactive): token auth", () => {
 
     const tempHome = await fs.mkdtemp(path.join(os.tmpdir(), "clawdbot-onboard-token-"));
     process.env.HOME = tempHome;
-    delete process.env.CLAWDBOT_STATE_DIR;
-    delete process.env.CLAWDBOT_CONFIG_PATH;
+    process.env.CLAWDBOT_STATE_DIR = tempHome;
+    process.env.CLAWDBOT_CONFIG_PATH = path.join(tempHome, "clawdbot.json");
+    vi.resetModules();
 
     const token = `sk-ant-oat01-${"a".repeat(80)}`;
 
diff --git a/src/memory/manager.atomic-reindex.test.ts b/src/memory/manager.atomic-reindex.test.ts
index 801bdba4b..a26b63a63 100644
--- a/src/memory/manager.atomic-reindex.test.ts
+++ b/src/memory/manager.atomic-reindex.test.ts
@@ -66,6 +66,7 @@ describe("memory manager atomic reindex", () => {
             provider: "openai",
             model: "mock-embed",
             store: { path: indexPath },
+            cache: { enabled: false },
             sync: { watch: false, onSessionStart: false, onSearch: false },
           },
         },
diff --git a/src/memory/manager.ts b/src/memory/manager.ts
index e5cb9c1ca..0e463d4d4 100644
--- a/src/memory/manager.ts
+++ b/src/memory/manager.ts
@@ -897,6 +897,62 @@ export class MemoryIndexManager {
     return path.join(dir, `${path.basename(basePath)}.tmp-${stamp}`);
   }
 
+  private seedEmbeddingCacheFrom(source: DatabaseSync): void {
+    if (!this.cache.enabled) return;
+    try {
+      const insert = this.db.prepare(
+        `INSERT INTO ${EMBEDDING_CACHE_TABLE} (provider, model, provider_key, hash, embedding, dims, updated_at)\n` +
+          ` VALUES (?, ?, ?, ?, ?, ?, ?)\n` +
+          ` ON CONFLICT(provider, model, provider_key, hash) DO UPDATE SET\n` +
+          `   embedding=excluded.embedding,\n` +
+          `   dims=excluded.dims,\n` +
+          `   updated_at=excluded.updated_at`,
+      );
+      const select = source.prepare(
+        `SELECT rowid, provider, model, provider_key, hash, embedding, dims, updated_at\n` +
+          ` FROM ${EMBEDDING_CACHE_TABLE}\n` +
+          ` WHERE provider = ? AND model = ? AND provider_key = ? AND rowid > ?\n` +
+          ` ORDER BY rowid\n` +
+          ` LIMIT ?`,
+      );
+      const batchSize = 500;
+      let lastRowId = 0;
+      while (true) {
+        const rows = select.all(
+          this.provider.id,
+          this.provider.model,
+          this.providerKey,
+          lastRowId,
+          batchSize,
+        ) as Array<{
+          rowid: number;
+          provider: string;
+          model: string;
+          provider_key: string;
+          hash: string;
+          embedding: string;
+          dims: number | null;
+          updated_at: number;
+        }>;
+        if (rows.length === 0) break;
+        for (const row of rows) {
+          insert.run(
+            row.provider,
+            row.model,
+            row.provider_key,
+            row.hash,
+            row.embedding,
+            row.dims,
+            row.updated_at,
+          );
+          lastRowId = row.rowid;
+        }
+      }
+    } catch {
+      // Swallow cache seed errors to avoid blocking indexing.
+    }
+  }
+
   private reopenDatabase() {
     this.db = this.openDatabase();
     this.fts.available = false;
@@ -962,6 +1018,7 @@ export class MemoryIndexManager {
     const tempPath = this.buildTempIndexPath();
     const scratch = this.createScratchManager(tempPath);
     try {
+      scratch.seedEmbeddingCacheFrom(this.db);
       await scratch.sync({ reason: params.reason, force: true, progress: params.progress });
     } catch (err) {
       await fs.rm(tempPath, { force: true });