feat: Add validation and improve documentation for embedder_config

- Add validation for embedder_config in Agent class
- Add test cases for invalid embedder configurations
- Improve docstrings with examples and error cases

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

src/crewai/agent.py

@@ -117,9 +117,29 @@ class Agent(BaseAgent):
     )
     embedder_config: Optional[Dict[str, Any]] = Field(
         default=None,
-        description="Embedder configuration for the agent.",
+        description="Embedder configuration for the agent. Must include 'provider' and relevant configuration parameters.",
     )
 
+    @validator("embedder_config")
+    def validate_embedder_config(cls, v):
+        """Validate embedder configuration.
+        
+        Args:
+            v: The embedder configuration to validate.
+            
+        Returns:
+            The validated embedder configuration.
+            
+        Raises:
+            ValueError: If the embedder configuration is invalid.
+        """
+        if v is not None:
+            if not isinstance(v, dict) or "provider" not in v:
+                raise ValueError("embedder_config must contain 'provider' key")
+            if "config" not in v:
+                raise ValueError("embedder_config must contain 'config' key")
+        return v
+
     @model_validator(mode="after")
     def post_init_setup(self):
         self._set_knowledge()

src/crewai/knowledge/storage/knowledge_storage.py

@@ -190,8 +190,18 @@ class KnowledgeStorage(BaseKnowledgeStorage):
         """Set the embedding configuration for the knowledge storage.
 
         Args:
-            embedder_config (Optional[Dict[str, Any]]): Configuration dictionary for the embedder.
+            embedder_config: Must include 'provider' and relevant configuration parameters.
-                If None or empty, defaults to the default embedding function.
+                For example:
+                {
+                    'provider': 'openai',
+                    'config': {
+                        'api_key': 'your-key',
+                        'model': 'text-embedding-3-small'
+                    }
+                }
+
+        Raises:
+            ValueError: If no configuration is provided or if the configuration is invalid.
         """
         self.embedder = (
             EmbeddingConfigurator().configure_embedder(embedder_config)

tests/test_agent_knowledge.py

@@ -36,6 +36,26 @@ def mock_vector_db():
         instance.reset.return_value = None
         yield instance
 
+def test_agent_invalid_embedder_config():
+    """Test that an invalid embedder configuration raises a ValueError."""
+    with pytest.raises(ValueError, match="embedder_config must contain 'provider' key"):
+        Agent(
+            role="test role",
+            goal="test goal",
+            backstory="test backstory",
+            knowledge_sources=[StringKnowledgeSource(content="test content")],
+            embedder_config={"invalid": "config"}
+        )
+
+    with pytest.raises(ValueError, match="embedder_config must contain 'config' key"):
+        Agent(
+            role="test role",
+            goal="test goal",
+            backstory="test backstory",
+            knowledge_sources=[StringKnowledgeSource(content="test content")],
+            embedder_config={"provider": "custom"}
+        )
+
 def test_agent_knowledge_with_custom_embedder(mock_vector_db):
     agent = Agent(
         role="test role",