Fix naming of the LLM providers

hnavidan · hnavidan · commit d7cdd9b92640 · 2025-08-04T20:07:56.000+02:00
diff --git a/backend/main.py b/backend/main.py
@@ -19,20 +19,20 @@
 
 
 PROVIDER_ENV_VAR_MAP = {
-    "google": "GOOGLE_API_KEY",
-    "anthropic": "ANTHROPIC_API_KEY",
-    "chatgpt": "OPENAI_API_KEY",
-    "grok": "XAI_API_KEY",
-    "glama": "GLAMA_API_KEY",
+    "Google": "GOOGLE_API_KEY",
+    "Anthropic": "ANTHROPIC_API_KEY",
+    "OpenAI": "OPENAI_API_KEY",
+    "xAI": "XAI_API_KEY",
+    "Glama": "GLAMA_API_KEY",
 }
 
 PROVIDER_API_CONFIG = {
-    "google": {
+    "Google": {
         "url_template": "https://generativelanguage.googleapis.com/v1beta/models?key={api_key}",
         "method": "GET",
         "auth_type": "query_param", # Key goes in URL
     },
-    "anthropic": {
+    "Anthropic": {
         "url": "https://api.anthropic.com/v1/models",
         "method": "GET",
         "auth_type": "header",
@@ -42,17 +42,17 @@
             "Content-Type": "application/json",
         }
     },
-    "chatgpt": { # OpenAI
+    "OpenAI": { # OpenAI
         "url": "https://api.openai.com/v1/models",
         "method": "GET",
         "auth_type": "bearer", # Authorization: Bearer {api_key}
     },
-    "grok": { # xAI
+    "xAI": { # xAI
         "url": "https://api.x.ai/v1/models", # Updated to correct xAI URL
         "method": "GET",
         "auth_type": "bearer",
     },
-     "glama": {
+     "Glama": {
         "url": "https://glama.ai/api/gateway/openai/v1/models", # Glama's specific endpoint
         "method": "GET",
         "auth_type": "bearer",
@@ -95,7 +95,7 @@ def fetch_and_format_models(provider: str, api_key: str) -> List[Dict[str, str]]
 
         # Parse the response based on provider structure
         models_data = []
-        if provider == "google":
+        if provider == "Google":
             models_data = data.get("models", [])
             # Filter and map Google models
             formatted_models = [
@@ -106,7 +106,7 @@ def fetch_and_format_models(provider: str, api_key: str) -> List[Dict[str, str]]
                 for m in models_data
                 if m.get("name") and "generateContent" in m.get("supportedGenerationMethods", [])
             ]
-        elif provider == "anthropic":
+        elif provider == "Anthropic":
             models_data = data.get("data", [])
             # Map Anthropic models
             formatted_models = [
@@ -189,26 +189,26 @@ def get_models():
 
 def create_llm_provider(provider_name: str, api_key: str):
     """Factory function to create LLM provider instances."""
-    if provider_name == "google":
+    if provider_name == "Google":
         if not GeminiProvider.validate_api_key(api_key):
             raise ValueError("Invalid Google API Key")
         return GeminiProvider(api_key)
-    elif provider_name == "chatgpt": # Assuming maps to OpenAIProvider
+    elif provider_name == "OpenAI": # Assuming maps to OpenAIProvider
         # Make sure OpenAIProvider exists and follows the pattern
         if not OpenAIProvider.validate_api_key(api_key):
             raise ValueError("Invalid OpenAI API Key")
         return OpenAIProvider(api_key)
-    elif provider_name == "grok":
+    elif provider_name == "xAI":
          # Make sure XAIProvider exists and follows the pattern
         if not XAIProvider.validate_api_key(api_key):
             raise ValueError("Invalid xAI API Key")
         return XAIProvider(api_key)
-    elif provider_name == "anthropic":
+    elif provider_name == "Anthropic":
          # Make sure AnthropicProvider exists and follows the pattern
         if not AnthropicProvider.validate_api_key(api_key):
             raise ValueError("Invalid Anthropic API Key")
         return AnthropicProvider(api_key)
-    elif provider_name == "glama": # <-- Add Glama case
+    elif provider_name == "Glama": # <-- Add Glama case
         if not GlamaProvider.validate_api_key(api_key):
             raise ValueError("Invalid Glama API Key")
         return GlamaProvider(api_key)
@@ -243,7 +243,7 @@ def chat():
 
         # Handle potential screenshot (currently only for Google Gemini)
         image_part = None
-        if screenshot_base64 and provider_name == "google":
+        if screenshot_base64 and provider_name == "Google":
             try:
                 # Ensure correct padding for base64
                 screenshot_data = screenshot_base64.split(',', 1)[1]
@@ -265,13 +265,13 @@ def chat():
                  return jsonify({"error": f"Invalid message format at index {i}: {message}"}), 400
 
             # Adjust roles if needed (e.g., Gemini specific adjustments)
-            if provider_name == "google":
+            if provider_name == "Google":
                 # Gemini uses 'user' and 'model'. Map 'system' to 'user'.
                 # The first message can often be 'system', treat it as 'user'.
                 # Subsequent system messages might need careful handling depending on context.
                 # Let's map system -> user for simplicity here.
                  target_role = 'user' if role in ['user', 'system'] else 'model'
-            elif provider_name == "glama" or provider_name == "chatgpt":
+            elif provider_name == "Glama" or provider_name == "OpenAI":
                  # OpenAI/Glama typically use 'user', 'assistant', 'system'
                  # Keep roles as they are, assuming frontend sends compatible roles
                  target_role = role
@@ -285,10 +285,10 @@ def chat():
             message_dict = {"role": target_role}
 
             # Add image part to the *last* message if it exists (Gemini logic)
-            if image_part and i == len(messages_input) - 1 and provider_name == "google":
+            if image_part and i == len(messages_input) - 1 and provider_name == "Google":
                  message_dict["parts"] = [{"text": content}, image_part]
             else:
-                 if provider_name == "google":
+                 if provider_name == "Google":
                      # Gemini needs parts structure
                      message_dict["parts"] = [{"text": content}]
                      # If image is present and it's the last message, add it
@@ -325,7 +325,7 @@ def chat():
 
         # --- Provider Instantiation and Call ---
         # Check for Glama required option *before* creating provider
-        if provider_name == "glama" and not options.get('model'):
+        if provider_name == "Glama" and not options.get('model'):
              return jsonify({"error": "Missing 'model' in options for Glama provider"}), 400
 
         llm_provider = create_llm_provider(provider_name, api_key)
diff --git a/provisioning/dashboards/dashboard.json b/provisioning/dashboards/dashboard.json
@@ -39,12 +39,12 @@
       "id": 1,
       "options": {
         "apiKey": "",
-        "backendAddr": "localhost",
+        "backendAddr": "http://localhost:5000",
         "context": "",
         "controlEndpointHeaders": "",
         "controlEndpointMethod": "POST",
         "controlEndpointUrl": "",
-        "llmProvider": "google",
+        "llmProvider": "Google",
         "model": ""
       },
       "pluginVersion": "1.0.0",
diff --git a/src/module.ts b/src/module.ts
@@ -11,7 +11,7 @@ const fetchModelsFromBackend = async (
 
   // Basic validation before calling backend
   if (!provider) { return [{ label: 'Select a provider', value: '' }]; }
-  if (!apiKey) { return [{ label: 'Enter API Key', value: '' }]; }
+  // if (!apiKey) { return [{ label: 'Enter API Key', value: '' }]; } No longer required as backend handles it
   if (!backendAddr) { return [{ label: 'Enter Backend Address', value: '', description: 'Required to fetch models' }]; }
 
   // Construct URL to your backend's new /models endpoint
@@ -71,34 +71,34 @@ const fetchModelsFromBackend = async (
 };
 
 export const plugin = new PanelPlugin<SimpleOptions>(LLMPanel).setPanelOptions((builder, context) => {
-  const providersWithModels = ['glama', 'google', 'chatgpt', 'grok', 'anthropic'];
+  const providersWithModels = ['Glama', 'Google', 'OpenAI', 'xAI', 'Anthropic'];
 
   return builder
     .addTextInput({
       path: 'backendAddr',
       name: 'Backend Address',
       description: 'Enter the address of the backend server. (e.g, http://localhost:5000)',
-      defaultValue: 'localhost', // Default to localhost
+      defaultValue: 'http://localhost:5000', 
     })
     .addSelect({
       path: 'llmProvider',
       name: 'API Provider',
       description: 'Select the API Provider',
       settings: {
         options: [
-          { label: 'Glama', value: 'glama' },
-          { label: 'xAI (Grok)', value: 'grok' },
-          { label: 'OpenAI', value: 'chatgpt' },
-          { label: 'Google', value: 'google' },
-          { label: 'Anthropic', value: 'anthropic' },
+          { label: 'Glama', value: 'Glama' },
+          { label: 'xAI (Grok)', value: 'xAI' },
+          { label: 'OpenAI', value: 'OpenAI' },
+          { label: 'Google', value: 'Google' },
+          { label: 'Anthropic', value: 'Anthropic' },
         ],
       },
-      defaultValue: 'google',
+      defaultValue: 'Google',
     })
     .addTextInput({
       path: 'apiKey',
-      name: 'API Key',
-      description: 'Enter your API key',
+      name: 'API Key (Optinoal)',
+      description: 'Leave blank to use the environment variables in the backend (Recommended). Keys entered here are not secure and are accessible by other users of the dashboard. Only use for testing purposes.',
       defaultValue: '',
       settings: {
         secure: true,
