fix: Remove OpenAI dependency for memory reset when using alternative LLMs

- Add environment variables for default embedding provider
- Support Ollama as default embedding provider
- Add tests for memory reset with different providers
- Update documentation

Fixes #2023

Co-Authored-By: Joe Moura <joao@crewai.com>

docs/memory.md

@@ -34,3 +34,43 @@ crewai reset-memories -a
 ```
 
 The memory system will use the configured embedding provider for all operations, including memory reset.
+
+## Troubleshooting
+
+### Common Issues
+
+1. OpenAI API Key Missing
+```bash
+# Error: EmbeddingConfigurationError: Invalid configuration for OpenAI provider
+# Solution: Set your OpenAI API key
+export OPENAI_API_KEY=your_key
+```
+
+2. Ollama Connection Issues
+```bash
+# Error: Failed to connect to Ollama server
+# Solution: Ensure Ollama is running and accessible
+curl http://localhost:11434/api/embeddings
+```
+
+3. Memory Reset Permission Issues
+```bash
+# Error: Failed to reset memory (readonly database)
+# Solution: Check file permissions or use custom path
+export CREWAI_MEMORY_PATH=/path/with/write/access
+```
+
+### Custom Storage Path
+
+You can configure a custom storage path for memory files:
+
+```python
+from crewai.memory import ShortTermMemory
+
+memory = ShortTermMemory(path="/custom/storage/path")
+```
+
+This is useful when:
+- Default storage location has permission issues
+- You want to isolate memory storage for different crews
+- You need to manage memory persistence manually

src/crewai/memory/storage/rag_storage.py

@@ -85,11 +85,10 @@ class RAGStorage(BaseRAGStorage):
 
         self._set_embedder_config()
         try:
-            chroma_client = chromadb.PersistentClient(
+            self.app = chromadb.PersistentClient(
                 path=self.path if self.path else self.storage_file_name,
                 settings=Settings(allow_reset=self.allow_reset),
             )
-            self.app = chroma_client
             if not self.app:
                 raise RuntimeError("Failed to initialize ChromaDB client")
 
@@ -104,19 +103,6 @@ class RAGStorage(BaseRAGStorage):
         except Exception as e:
             raise RuntimeError(f"Failed to initialize ChromaDB: {str(e)}")
 
-        self.app = chroma_client
-        if not self.app:
-            raise RuntimeError("Failed to initialize ChromaDB client")
-
-        try:
-            self.collection = self.app.get_collection(
-                name=self.type, embedding_function=self.embedder_config
-            )
-        except Exception:
-            self.collection = self.app.create_collection(
-                name=self.type, embedding_function=self.embedder_config
-            )
-
     def _sanitize_role(self, role: str) -> str:
         """
         Sanitizes agent roles to ensure valid directory names.
@@ -138,12 +124,21 @@ class RAGStorage(BaseRAGStorage):
         return f"{base_path}/{file_name}"
 
     def save(self, value: Any, metadata: Dict[str, Any]) -> None:
+        """Save a value with metadata to the memory storage.
+
+        Args:
+            value: The text content to store
+            metadata: Additional metadata for the stored content
+
+        Raises:
+            EmbeddingInitializationError: If embedding generation fails
+        """
         if not hasattr(self, "app") or not hasattr(self, "collection"):
             self._initialize_app()
         try:
             self._generate_embedding(value, metadata)
         except Exception as e:
-            logging.error(f"Error during {self.type} save: {str(e)}")
+            raise EmbeddingInitializationError(self.type, str(e))
 
     def search(
         self,
@@ -151,7 +146,18 @@ class RAGStorage(BaseRAGStorage):
         limit: int = 3,
         filter: Optional[dict] = None,
         score_threshold: float = 0.35,
-    ) -> List[Any]:
+    ) -> List[Dict[str, Any]]:
+        """Search for similar content in memory.
+
+        Args:
+            query: The search query text
+            limit: Maximum number of results to return
+            filter: Optional filter criteria
+            score_threshold: Minimum similarity score threshold
+
+        Returns:
+            List of matching results with metadata and scores
+        """
         if not hasattr(self, "app"):
             self._initialize_app()
 
@@ -175,15 +181,24 @@ class RAGStorage(BaseRAGStorage):
             logging.error(f"Error during {self.type} search: {str(e)}")
             return []
 
-    def _generate_embedding(self, text: str, metadata: Dict[str, Any]) -> None:  # type: ignore
+    def _generate_embedding(self, text: str, metadata: Dict[str, Any]) -> None:
+        """Generate and store embeddings for the given text.
+
+        Args:
+            text: The text to generate embeddings for
+            metadata: Additional metadata to store with the embeddings
+        """
         if not hasattr(self, "app") or not hasattr(self, "collection"):
             self._initialize_app()
 
-        self.collection.add(
-            documents=[text],
-            metadatas=[metadata or {}],
-            ids=[str(uuid.uuid4())],
-        )
+        try:
+            self.collection.add(
+                documents=[text],
+                metadatas=[metadata or {}],
+                ids=[str(uuid.uuid4())],
+            )
+        except Exception as e:
+            raise EmbeddingInitializationError(self.type, f"Failed to generate embedding: {str(e)}")
 
     def reset(self) -> None:
         """Reset the memory storage by clearing the database and removing files.

src/crewai/utilities/embedding_configurator.py

@@ -71,8 +71,11 @@ class EmbeddingConfigurator:
         model = os.getenv("CREWAI_EMBEDDING_MODEL", DEFAULT_EMBEDDING_MODEL)
         
         if provider == "openai":
+            api_key = os.getenv("OPENAI_API_KEY")
+            if not api_key:
+                raise EmbeddingConfigurationError("OpenAI API key is required but not provided")
             from chromadb.utils.embedding_functions.openai_embedding_function import OpenAIEmbeddingFunction
-            return OpenAIEmbeddingFunction(api_key=os.getenv("OPENAI_API_KEY"), model_name=model)
+            return OpenAIEmbeddingFunction(api_key=api_key, model_name=model)
         elif provider == "ollama":
             from chromadb.utils.embedding_functions.ollama_embedding_function import OllamaEmbeddingFunction
             url = os.getenv("CREWAI_OLLAMA_URL", "http://localhost:11434/api/embeddings")

tests/memory/test_memory_reset.py

@@ -48,6 +48,19 @@ def test_memory_reset_with_invalid_provider(temp_db_dir):
         for memory in memories:
             memory.reset()
 
+def test_memory_reset_with_invalid_configuration(temp_db_dir):
+    os.environ["CREWAI_EMBEDDING_PROVIDER"] = "openai"
+    os.environ.pop("OPENAI_API_KEY", None)
+    
+    with pytest.raises(EmbeddingConfigurationError):
+        memories = [
+            ShortTermMemory(path=temp_db_dir),
+            LongTermMemory(path=temp_db_dir),
+            EntityMemory(path=temp_db_dir)
+        ]
+        for memory in memories:
+            memory.reset()
+
 def test_memory_reset_with_missing_ollama_url(temp_db_dir):
     os.environ["CREWAI_EMBEDDING_PROVIDER"] = "ollama"
     os.environ.pop("CREWAI_OLLAMA_URL", None)
@@ -59,3 +72,20 @@ def test_memory_reset_with_missing_ollama_url(temp_db_dir):
     ]
     for memory in memories:
         memory.reset()
+
+def test_memory_reset_with_custom_path(temp_db_dir):
+    os.environ["CREWAI_EMBEDDING_PROVIDER"] = "ollama"
+    custom_path = os.path.join(temp_db_dir, "custom")
+    os.makedirs(custom_path, exist_ok=True)
+    
+    memories = [
+        ShortTermMemory(path=custom_path),
+        LongTermMemory(path=custom_path),
+        EntityMemory(path=custom_path)
+    ]
+    for memory in memories:
+        memory.reset()
+        
+    assert not os.path.exists(os.path.join(custom_path, "short_term"))
+    assert not os.path.exists(os.path.join(custom_path, "long_term"))
+    assert not os.path.exists(os.path.join(custom_path, "entity"))