fix: cherry-pick improvements from feature/semantic_caching

- semantic-cache.ts: use upsert instead of add to avoid duplicate-ID errors
- semantic-cache.ts: add crypto.randomUUID() fallback for constrained runtimes
- ollama-client.ts: JSON.stringify non-string Ollama errors to avoid [object Object]
- tests: update semantic-cache mocks and assertions for upsert

src/ollama-client.ts

@@ -196,7 +196,9 @@ export class OllamaClient {
           }
 
           if (parsed.error) {
-            throw new Error(`Ollama error: ${parsed.error}`);
+            const errorMsg =
+              typeof parsed.error === 'string' ? parsed.error : JSON.stringify(parsed.error);
+            throw new Error(`Ollama error: ${errorMsg}`);
           }
 
           yield this.normalizeMessage(parsed.message);
@@ -216,7 +218,9 @@ export class OllamaClient {
         }
 
         if (parsed?.error) {
-          throw new Error(`Ollama error: ${parsed.error}`);
+          const errorMsg =
+            typeof parsed.error === 'string' ? parsed.error : JSON.stringify(parsed.error);
+          throw new Error(`Ollama error: ${errorMsg}`);
         }
 
         if (parsed?.message) {
@@ -226,7 +230,10 @@ export class OllamaClient {
     } catch (error) {
       if (retryCount < this.maxRetries && this.isRetryableError(error, controller)) {
         const errorMessage = error instanceof Error ? error.message : String(error);
-        Logger.warn(`Retrying after error (attempt ${retryCount + 1}): ${errorMessage}`, 'ollama-client');
+        Logger.warn(
+          `Retrying after error (attempt ${retryCount + 1}): ${errorMessage}`,
+          'ollama-client'
+        );
         await new Promise((resolve) => setTimeout(resolve, 10 * Math.pow(2, retryCount)));
         yield* this.streamChatWithRetry(messages, tools, retryCount + 1);
       } else {
@@ -267,7 +274,7 @@ export class OllamaClient {
         throw new ApiError(`Ollama API error: ${response.status}`, response.status);
       }
 
-      const data = await response.json() as unknown;
+      const data = (await response.json()) as unknown;
       if (!this.isChatResponse(data)) {
         return this.normalizeMessage();
       }
@@ -275,7 +282,10 @@ export class OllamaClient {
     } catch (error) {
       if (retryCount < this.maxRetries && this.isRetryableError(error, controller)) {
         const errorMessage = error instanceof Error ? error.message : String(error);
-        Logger.warn(`Retrying after error (attempt ${retryCount + 1}): ${errorMessage}`, 'ollama-client');
+        Logger.warn(
+          `Retrying after error (attempt ${retryCount + 1}): ${errorMessage}`,
+          'ollama-client'
+        );
         await new Promise((resolve) => setTimeout(resolve, 10 * Math.pow(2, retryCount)));
         return this.chatWithRetry(messages, tools, retryCount + 1);
       } else {

src/semantic-cache.ts

@@ -62,12 +62,20 @@ export class SemanticCacheService {
     }
   }
 
+  private static generateId(): string {
+    if (typeof crypto !== 'undefined' && typeof crypto.randomUUID === 'function') {
+      return crypto.randomUUID();
+    }
+    // Fallback for environments without crypto.randomUUID
+    return 'cache_' + Date.now() + '_' + Math.random().toString(36).substring(2, 11);
+  }
+
   async setCache(query: string, response: string): Promise<void> {
     if (!this.config.enabled || !this.collection) return;
 
     try {
-      await this.collection.add({
-        ids: [crypto.randomUUID()],
+      await this.collection.upsert({
+        ids: [SemanticCacheService.generateId()],
         documents: [response],
         embeddings: await this.generateEmbedding(query),
         metadatas: [{ source: 'ollama' }],

tests/semantic-cache.test.ts

@@ -10,6 +10,7 @@ jest.mock('chromadb', () => ({
       getOrCreateCollection: jest.fn().mockResolvedValue({
         query: jest.fn(),
         add: jest.fn(),
+        upsert: jest.fn(),
         reset: jest.fn(),
       }),
       deleteCollection: jest.fn(),
@@ -135,13 +136,13 @@ describe('SemanticCacheService', () => {
 
       await disabledCacheService.setCache('test query', 'test response');
 
-      expect(mockCollection.add).not.toHaveBeenCalled();
+      expect(mockCollection.upsert).not.toHaveBeenCalled();
     });
 
     it('should add content to cache', async () => {
       await cacheService.setCache('test query', 'test response');
 
-      expect(mockCollection.add).toHaveBeenCalled();
+      expect(mockCollection.upsert).toHaveBeenCalled();
     });
   });