feat: Add model name validation and expand test coverage

- Add validation for Azure and Cerebras model names
- Add validation handling in create_crew.py
- Expand test coverage for model env var cases

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

src/crewai/cli/constants.py

@@ -66,7 +66,8 @@ ENV_VARS = {
     "azure": [
         {
             "prompt": "Enter your Azure deployment name (must start with 'azure/')",
-            "key_name": "model",
+            "key_name": "MODEL",  # Uppercase MODEL used for consistency across environment variables
+            "validator": lambda x: x.startswith("azure/") or "Model name must start with 'azure/'"
         },
         {
             "prompt": "Enter your AZURE API key (press Enter to skip)",
@@ -84,7 +85,8 @@ ENV_VARS = {
     "cerebras": [
         {
             "prompt": "Enter your Cerebras model name (must start with 'cerebras/')",
-            "key_name": "model",
+            "key_name": "MODEL",  # Uppercase MODEL used for consistency across environment variables
+            "validator": lambda x: x.startswith("cerebras/") or "Model name must start with 'cerebras/'"
         },
         {
             "prompt": "Enter your Cerebras API version (press Enter to skip)",

src/crewai/cli/create_crew.py

@@ -157,10 +157,19 @@ def create_crew(name, provider=None, skip_provider=False, parent_folder=None):
                     # Prompt for non-default key-value pairs
                     prompt = details["prompt"]
                     key_name = details["key_name"]
-                    api_key_value = click.prompt(prompt, default="", show_default=False)
+                    while True:
+                        api_key_value = click.prompt(prompt, default="", show_default=False)
+                        if not api_key_value.strip():
+                            break
+
+                        if "validator" in details:
+                            validation_result = details["validator"](api_key_value)
+                            if isinstance(validation_result, str):
+                                click.secho(f"Invalid input: {validation_result}", fg="red")
+                                continue
 
-                    if api_key_value.strip():
                         env_vars[key_name] = api_key_value
+                        break
 
         if env_vars:
             write_env_file(folder_path, env_vars)

tests/agent_test.py

@@ -21,6 +21,38 @@ from crewai.utilities import RPMController
 from crewai.utilities.events import Emitter
 
 
+def test_agent_model_env_var():
+    """Test MODEL environment variable handling with various cases."""
+    # Store original environment variables
+    original_model = os.environ.get("MODEL")
+
+    test_cases = [
+        ("azure/test-model", "azure/test-model"),  # Valid Azure case
+        ("azure/minimal", "azure/minimal"),  # Another valid Azure case
+        ("cerebras/test-model", "cerebras/test-model"),  # Valid Cerebras case
+        ("cerebras/minimal", "cerebras/minimal"),  # Another valid Cerebras case
+    ]
+
+    for input_model, expected_model in test_cases:
+        # Set test MODEL value
+        os.environ["MODEL"] = input_model
+        agent = Agent(role="test role", goal="test goal", backstory="test backstory")
+        assert agent.llm.model == expected_model
+
+    # Test missing MODEL env var
+    if "MODEL" in os.environ:
+        del os.environ["MODEL"]
+    agent = Agent(role="test role", goal="test goal", backstory="test backstory")
+    assert agent.llm.model == "gpt-4o-mini"  # Default model
+
+    # Clean up environment variables
+    if original_model:
+        os.environ["MODEL"] = original_model
+    else:
+        if "MODEL" in os.environ:
+            del os.environ["MODEL"]
+
+
 def test_agent_llm_creation_with_env_vars():
     # Store original environment variables
     original_api_key = os.environ.get("OPENAI_API_KEY")