Add user-specific API endpoints and MCP configuration

- Added /api/user/* endpoints for authenticated users to access their own configurations
- Admin users get access to global + user configs, regular users get user-only
- Global tokens from aisbf.json have full access to all configurations
- Enhanced MCP with user-specific tools for authenticated users
- Updated user dashboard with comprehensive API endpoint documentation
- Updated README.md, DOCUMENTATION.md with new endpoint documentation
- Updated CHANGELOG.md with new features
- Bumped version to 0.9.1

CHANGELOG.md

@@ -2,6 +2,21 @@
 
 ## [Unreleased]
 ### Added
+- **User-Specific API Endpoints**: New API endpoints for authenticated users to access their own configurations
+  - `GET /api/user/models` - List user's own models
+  - `GET /api/user/providers` - List user's provider configurations  
+  - `GET /api/user/rotations` - List user's rotation configurations
+  - `GET /api/user/autoselects` - List user's autoselect configurations
+  - `POST /api/user/chat/completions` - Chat completions using user's own models
+  - `GET /api/user/{config_type}/models` - List models for specific config type
+  - Requires Bearer token or query parameter authentication
+  - Admin users get access to global + user configs, regular users get user-only configs
+  - Global tokens (in aisbf.json) have full access to all configurations
+- **MCP User Configuration**: Enhanced MCP server with user-specific tools for authenticated users
+  - User can configure their own models, providers, autoselects, and rotations through MCP
+  - Admin users get access to both global and user tools
+  - Regular users get access to user-only tools
+- **Dashboard API Documentation**: User dashboard now includes comprehensive API endpoint documentation
 - **Adaptive Rate Limiting**: Intelligent rate limit management that learns from 429 responses
   - Per-provider adaptive rate limiters with learning capability
   - Exponential backoff with jitter (configurable base and jitter factor)

DOCUMENTATION.md

@@ -563,6 +563,68 @@ Cache backend can be configured via the web dashboard or configuration file:
   - Supports both streaming and non-streaming responses
 - `GET /api/autoselect/models` - List all models across all autoselect configurations
 
+### User-Specific API Endpoints
+
+Authenticated users can access their own configurations via user-specific API endpoints. These endpoints require either a valid API token (generated in the user dashboard) or session authentication.
+
+#### Authentication
+
+**Option 1: Bearer Token (Recommended for API access)**
+```bash
+Authorization: Bearer YOUR_API_TOKEN
+```
+
+**Option 2: Query Parameter**
+```bash
+?token=YOUR_API_TOKEN
+```
+
+#### User API Endpoints
+
+| Endpoint | Description |
+|----------|-------------|
+| `GET /api/user/models` | List available models from user's own configurations |
+| `GET /api/user/providers` | List user's provider configurations |
+| `GET /api/user/rotations` | List user's rotation configurations |
+| `GET /api/user/autoselects` | List user's autoselect configurations |
+| `POST /api/user/chat/completions` | Chat completions using user's own models |
+| `GET /api/user/{config_type}/models` | List models for specific config type (provider, rotation, autoselect) |
+
+**Access Control:**
+- **Admin Users** have access to both global and user configurations when using user API endpoints
+- **Regular Users** can only access their own configurations
+- **Global Tokens** (configured in aisbf.json) have full access to all configurations
+
+#### Example: Using User API with cURL
+
+```bash
+# List user models
+curl -H "Authorization: Bearer YOUR_TOKEN" http://localhost:17765/api/user/models
+
+# Chat using user's own models
+curl -X POST -H "Authorization: Bearer YOUR_TOKEN" \
+  -H "Content-Type: application/json" \
+  -d '{"model": "your-rotation/model", "messages": [{"role": "user", "content": "Hello"}]}' \
+  http://localhost:17765/api/user/chat/completions
+```
+
+### MCP (Model Context Protocol)
+
+AISBF provides an MCP server for remote agent configuration and model access:
+
+- **SSE Endpoint**: `GET /mcp` - Server-Sent Events for MCP communication
+- **HTTP Endpoint**: `POST /mcp` - Direct HTTP transport for MCP
+
+MCP tools include:
+- `list_models` - List available models (user or global depending on auth)
+- `chat_completions` - Send chat completion requests
+- `get_providers` - Get provider configurations
+- `get_rotations` - Get rotation configurations
+- `get_autoselects` - Get autoselect configurations
+- And more for authenticated users to manage their own configs
+
+User tokens authenticate MCP requests, with admin users getting full access and regular users getting user-only access.
+
 ## Provider Support
 
 AISBF supports the following AI providers:

README.md

@@ -648,6 +648,69 @@ These endpoints are maintained for backward compatibility:
 - `GET /api/autoselect/models` - List all models across all autoselect configurations
 - `GET /api/{provider_id}/models` - List available models for a specific provider, rotation, or autoselect
 
+### User-Specific API Endpoints
+
+Authenticated users can access their own configurations via user-specific API endpoints. These endpoints require either a valid API token (generated in the user dashboard) or session authentication.
+
+#### Authentication
+
+**Option 1: Bearer Token (Recommended for API access)**
+```bash
+Authorization: Bearer YOUR_API_TOKEN
+```
+
+**Option 2: Query Parameter**
+```bash
+?token=YOUR_API_TOKEN
+```
+
+#### User API Endpoints
+
+| Endpoint | Description |
+|----------|-------------|
+| `GET /api/user/models` | List available models from user's own configurations |
+| `GET /api/user/providers` | List user's provider configurations |
+| `GET /api/user/rotations` | List user's rotation configurations |
+| `GET /api/user/autoselects` | List user's autoselect configurations |
+| `POST /api/user/chat/completions` | Chat completions using user's own models |
+| `GET /api/user/{config_type}/models` | List models for specific config type (provider, rotation, autoselect) |
+
+**Admin Users** have access to both global and user configurations when using user API endpoints.
+
+**Regular Users** can only access their own configurations.
+
+**Global Tokens** (configured in aisbf.json) have full access to all configurations.
+
+#### Example: Using User API with cURL
+
+```bash
+# List user models
+curl -H "Authorization: Bearer YOUR_TOKEN" http://localhost:17765/api/user/models
+
+# Chat using user's own models
+curl -X POST -H "Authorization: Bearer YOUR_TOKEN" \
+  -H "Content-Type: application/json" \
+  -d '{"model": "your-rotation/model", "messages": [{"role": "user", "content": "Hello"}]}' \
+  http://localhost:17765/api/user/chat/completions
+```
+
+### MCP (Model Context Protocol)
+
+AISBF provides an MCP server for remote agent configuration and model access:
+
+- **SSE Endpoint**: `GET /mcp` - Server-Sent Events for MCP communication
+- **HTTP Endpoint**: `POST /mcp` - Direct HTTP transport for MCP
+
+MCP tools include:
+- `list_models` - List available models (user or global depending on auth)
+- `chat_completions` - Send chat completion requests
+- `get_providers` - Get provider configurations
+- `get_rotations` - Get rotation configurations
+- `get_autoselects` - Get autoselect configurations
+- And more for authenticated users to manage their own configs
+
+User tokens authenticate MCP requests, with admin users getting full access and regular users getting user-only access.
+
 ### Content Proxy
 - `GET /api/proxy/{content_id}` - Proxy generated content (images, audio, etc.)
 

aisbf/config.py

@@ -141,6 +141,16 @@ class AutoselectConfig(BaseModel):
     pricing: Optional[Dict] = None
     supported_parameters: Optional[List[str]] = None
     default_parameters: Optional[Dict] = None
+    # Default settings for models in this autoselect
+    default_rate_limit: Optional[float] = None
+    default_max_request_tokens: Optional[int] = None
+    default_rate_limit_TPM: Optional[int] = None
+    default_rate_limit_TPH: Optional[int] = None
+    default_rate_limit_TPD: Optional[int] = None
+    default_context_size: Optional[int] = None
+    default_condense_context: Optional[int] = None
+    default_condense_method: Optional[Union[str, List[str]]] = None
+    default_error_cooldown: Optional[int] = None  # Default cooldown period in seconds after 3 consecutive failures (default: 300)
     # Response caching control
     enable_response_cache: Optional[bool] = None  # Enable/disable response caching for this autoselect (None = use global default)
 

aisbf/context.py

@@ -971,9 +971,10 @@ def get_context_config_for_model(
     
     Priority order for each field:
     1. Rotation model config (if explicitly set)
-    2. Provider model-specific config (if exists)
-    3. Provider default config (fallback)
-    4. System default (0 for condense_context, None for others)
+    2. Model-specific config in provider (if exists)
+    3. First model in provider (auto-derived from dynamic fetch)
+    4. Provider default config (fallback)
+    5. System default (0 for condense_context, None for others)
     
     Args:
         model_name: Name of the model
@@ -1007,12 +1008,23 @@ def get_context_config_for_model(
                         model_specific_config = model
                     break
         
-        # Build base config from provider (model-specific > provider defaults)
+        # Build base config from provider (model-specific > first model > provider defaults)
         # context_size
         if model_specific_config and model_specific_config.get('context_size') is not None:
             context_config['context_size'] = model_specific_config.get('context_size')
         elif hasattr(provider_config, 'default_context_size') and provider_config.default_context_size is not None:
             context_config['context_size'] = provider_config.default_context_size
+        else:
+            # Auto-derive from first model in provider (has context_size from dynamic fetch)
+            if provider_config.models and len(provider_config.models) > 0:
+                first_model = provider_config.models[0]
+                # Check for context_size in the first model (from dynamic fetch)
+                if hasattr(first_model, 'context_size') and first_model.context_size:
+                    context_config['context_size'] = first_model.context_size
+                elif hasattr(first_model, 'context_window') and first_model.context_window:
+                    context_config['context_size'] = first_model.context_window
+                elif hasattr(first_model, 'context_length') and first_model.context_length:
+                    context_config['context_size'] = first_model.context_length
         
         # condense_context
         if model_specific_config and model_specific_config.get('condense_context') is not None:
@@ -1033,11 +1045,69 @@ def get_context_config_for_model(
         # Only override if the field is explicitly set in rotation config
         if 'context_size' in rotation_model_config and rotation_model_config['context_size'] is not None:
             context_config['context_size'] = rotation_model_config['context_size']
+        elif context_config.get('context_size') is None:
+            # If context_size is still None, check if rotation_model_config has default_context_size
+            # (This handles the case where rotation-level default should be used)
+            if 'default_context_size' in rotation_model_config and rotation_model_config['default_context_size'] is not None:
+                context_config['context_size'] = rotation_model_config['default_context_size']
         
         if 'condense_context' in rotation_model_config and rotation_model_config['condense_context'] is not None:
             context_config['condense_context'] = rotation_model_config['condense_context']
+        elif context_config.get('condense_context') == 0:
+            # If condense_context is still at default (0), check if rotation has default
+            if 'default_condense_context' in rotation_model_config and rotation_model_config['default_condense_context'] is not None:
+                context_config['condense_context'] = rotation_model_config['default_condense_context']
         
         if 'condense_method' in rotation_model_config and rotation_model_config['condense_method'] is not None:
             context_config['condense_method'] = rotation_model_config['condense_method']
     
-    return context_config
+    # Step 3: Final fallback if context_size is still None
\ No newline at end of file
+    # Use inference based on model name patterns
+    if context_config.get('context_size') is None:
+        context_config['context_size'] = _infer_context_size_from_model(model_name)
+    
+    return context_config
+
+
+def _infer_context_size_from_model(model_name: str) -> int:
+    """
+    Infer context window size from model name patterns.
+    
+    Args:
+        model_name: Name of the model
+        
+    Returns:
+        Inferred context size in tokens
+    """
+    model_lower = model_name.lower()
+    
+    # Known model patterns
+    if 'gpt-4' in model_lower:
+        if 'turbo' in model_lower or '1106' in model_lower or '0125' in model_lower:
+            return 128000
+        return 8192
+    elif 'gpt-3.5' in model_lower:
+        if 'turbo' in model_lower and ('1106' in model_lower or '0125' in model_lower):
+            return 16385
+        return 4096
+    elif 'claude-3' in model_lower:
+        return 200000
+    elif 'claude-2' in model_lower:
+        return 100000
+    elif 'gemini' in model_lower:
+        if '1.5' in model_lower:
+            return 2000000 if 'pro' in model_lower else 1000000
+        elif '2.0' in model_lower:
+            return 1000000
+        return 32000
+    elif 'llama' in model_lower:
+        if '3' in model_lower:
+            return 128000
+        return 4096
+    elif 'mistral' in model_lower:
+        if 'large' in model_lower:
+            return 32000
+        return 8192
+    
+    # Generic default
+    return 8192
\ No newline at end of file

aisbf/handlers.py

@@ -1505,7 +1505,8 @@ class RotationHandler:
         Priority order:
         1. Model-specific settings (highest priority)
         2. Rotation default settings
-        3. Provider default settings (lowest priority)
+        3. Provider default settings
+        4. Auto-derived from first model in provider (lowest priority)
         
         Args:
             model: The model configuration dict
@@ -1539,9 +1540,108 @@ class RotationHandler:
                     provider_default = getattr(provider_config, f'default_{field}', None)
                     if provider_default is not None:
                         model[field] = provider_default
+                    else:
+                        # Auto-derive from first model in provider config if available
+                        if provider_config and provider_config.models and len(provider_config.models) > 0:
+                            first_model = provider_config.models[0]
+                            # For context_size, check multiple field names (from dynamic fetch)
+                            if field == 'context_size':
+                                model_field = getattr(first_model, 'context_size', None)
+                                if model_field is None:
+                                    model_field = getattr(first_model, 'context_window', None)
+                                if model_field is None:
+                                    model_field = getattr(first_model, 'context_length', None)
+                            else:
+                                model_field = getattr(first_model, field, None)
+                            if model_field is not None:
+                                model[field] = model_field
         
         return model
 
+    def _apply_defaults_to_autoselect_model(self, model_config: Dict, autoselect_config) -> Dict:
+        """
+        Apply default settings to an autoselect model configuration.
+        
+        Priority order:
+        1. Model-specific settings (highest priority)
+        2. Autoselect default settings
+        3. Auto-derived from first model in rotation (lowest priority)
+        
+        Args:
+            model_config: The model configuration dict (typically a rotation_id from autoselect)
+            autoselect_config: The autoselect configuration
+            
+        Returns:
+            Model dict with defaults applied
+        """
+        import logging
+        logger = logging.getLogger(__name__)
+        
+        # List of fields that can have defaults
+        default_fields = [
+            'rate_limit',
+            'max_request_tokens',
+            'rate_limit_TPM',
+            'rate_limit_TPH',
+            'rate_limit_TPD',
+            'context_size',
+            'condense_context',
+            'condense_method'
+        ]
+        
+        # First, check if the model_config is a rotation ID and get its settings
+        model_id = model_config.get('model_id') or model_config.get('name') or model_config.get('id', '')
+        
+        # Try to get defaults from the referenced rotation (first model in the rotation)
+        if model_id in self.config.rotations:
+            rotation_config = self.config.rotations[model_id]
+            
+            # Check each default field
+            for field in default_fields:
+                # If field is not set in model, try autoselect defaults, then rotation defaults
+                if field not in model_config or model_config.get(field) is None:
+                    # Try autoselect defaults first
+                    autoselect_default = getattr(autoselect_config, f'default_{field}', None)
+                    if autoselect_default is not None:
+                        model_config[field] = autoselect_default
+                        logger.debug(f"Applied autoselect default_{field}: {autoselect_default} to model {model_id}")
+                    else:
+                        # Try rotation defaults
+                        rotation_default = getattr(rotation_config, f'default_{field}', None)
+                        if rotation_default is not None:
+                            model_config[field] = rotation_default
+                            logger.debug(f"Applied rotation default_{field}: {rotation_default} to model {model_id}")
+                        else:
+                            # Auto-derive from first provider in rotation, then first model
+                            if rotation_config.providers and len(rotation_config.providers) > 0:
+                                first_provider = rotation_config.providers[0]
+                                provider_id = first_provider.get('provider_id')
+                                provider_config = self.config.get_provider(provider_id)
+                                
+                                if provider_config and provider_config.models and len(provider_config.models) > 0:
+                                    first_model = provider_config.models[0]
+                                    # Check for context_size, context_window, or context_length
+                                    if field == 'context_size':
+                                        model_field = getattr(first_model, 'context_size', None)
+                                        if model_field is None:
+                                            model_field = getattr(first_model, 'context_window', None)
+                                        if model_field is None:
+                                            model_field = getattr(first_model, 'context_length', None)
+                                    else:
+                                        model_field = getattr(first_model, field, None)
+                                    if model_field is not None:
+                                        model_config[field] = model_field
+                                        logger.debug(f"Auto-derived default_{field}: {model_field} from first model in {provider_id}")
+        else:
+            # Not a rotation, apply autoselect defaults directly
+            for field in default_fields:
+                if field not in model_config or model_config.get(field) is None:
+                    autoselect_default = getattr(autoselect_config, f'default_{field}', None)
+                    if autoselect_default is not None:
+                        model_config[field] = autoselect_default
+        
+        return model_config
+
     async def _handle_chunked_rotation_request(
         self,
         handler,
@@ -2994,16 +3094,31 @@ class RotationHandler:
                 }
                 
                 # Add context window information
+                # Priority: model config in rotation > provider config > first model in provider
                 if model.get('context_size'):
                     model_dict['context_window'] = model['context_size']
                 elif provider_config:
                     # Try to find in provider config
+                    found_in_provider = False
                     for pm in provider_config.models or []:
-                        if pm.name == model_name and hasattr(pm, 'context_size'):
+                        if pm.name == model_name and hasattr(pm, 'context_size') and pm.context_size:
                             model_dict['context_window'] = pm.context_size
+                            found_in_provider = True
                             break
-                    if 'context_window' not in model_dict:
-                        model_dict['context_window'] = self._infer_context_window(model_name, provider_config.type)
+                    if not found_in_provider:
+                        # Auto-derive from first model in provider (which has context_size from dynamic fetch)
+                        if provider_config.models and len(provider_config.models) > 0:
+                            first_model = provider_config.models[0]
+                            if hasattr(first_model, 'context_size') and first_model.context_size:
+                                model_dict['context_window'] = first_model.context_size
+                            elif hasattr(first_model, 'context_window') and first_model.context_window:
+                                model_dict['context_window'] = first_model.context_window
+                            elif hasattr(first_model, 'context_length') and first_model.context_length:
+                                model_dict['context_window'] = first_model.context_length
+                            else:
+                                model_dict['context_window'] = self._infer_context_window(model_name, provider_config.type)
+                        else:
+                            model_dict['context_window'] = self._infer_context_window(model_name, provider_config.type)
                 
                 # Add capabilities information
                 if model.get('capabilities'):

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "aisbf"
-version = "0.9.0"
+version = "0.9.1"
 description = "AISBF - AI Service Broker Framework || AI Should Be Free - A modular proxy server for managing multiple AI provider integrations"
 readme = "README.md"
 license = "GPL-3.0-or-later"

setup.py

@@ -49,7 +49,7 @@ class InstallCommand(_install):
 
 setup(
     name="aisbf",
-    version="0.9.0",
+    version="0.9.1",
     author="AISBF Contributors",
     author_email="stefy@nexlab.net",
     description="AISBF - AI Service Broker Framework || AI Should Be Free - A modular proxy server for managing multiple AI provider integrations",

templates/dashboard/autoselect.html

@@ -197,6 +197,50 @@ function renderAutoselectDetails(autoselectKey) {
             <small style="color: #a0a0a0; font-size: 12px; display: block; margin-top: 5px;">Choose from rotations or provider models</small>
         </div>
         
+        <h4 style="margin-top: 20px; margin-bottom: 10px;">Default Settings</h4>
+        <p style="color: #a0a0a0; font-size: 14px; margin-bottom: 10px;">Default values for models in this autoselect (optional - auto-derived from first model if not set)</p>
+        
+        <div style="display: grid; grid-template-columns: 1fr 1fr; gap: 15px;">
+            <div class="form-group">
+                <label>Default Rate Limit (seconds)</label>
+                <input type="number" value="${autoselect.default_rate_limit || ''}" onchange="updateAutoselect('${autoselectKey}', 'default_rate_limit', this.value ? parseFloat(this.value) : null)" step="0.1" placeholder="Optional">
+            </div>
+            
+            <div class="form-group">
+                <label>Default Max Request Tokens</label>
+                <input type="number" value="${autoselect.default_max_request_tokens || ''}" onchange="updateAutoselect('${autoselectKey}', 'default_max_request_tokens', this.value ? parseInt(this.value) : null)" placeholder="Optional">
+            </div>
+            
+            <div class="form-group">
+                <label>Default Context Size</label>
+                <input type="number" value="${autoselect.default_context_size || ''}" onchange="updateAutoselect('${autoselectKey}', 'default_context_size', this.value ? parseInt(this.value) : null)" placeholder="Optional">
+            </div>
+            
+            <div class="form-group">
+                <label>Default Rate Limit TPM</label>
+                <input type="number" value="${autoselect.default_rate_limit_TPM || ''}" onchange="updateAutoselect('${autoselectKey}', 'default_rate_limit_TPM', this.value ? parseInt(this.value) : null)" placeholder="Optional">
+                <small style="color: #a0a0a0; font-size: 12px; display: block; margin-top: 5px;">Tokens per minute limit</small>
+            </div>
+            
+            <div class="form-group">
+                <label>Default Rate Limit TPH</label>
+                <input type="number" value="${autoselect.default_rate_limit_TPH || ''}" onchange="updateAutoselect('${autoselectKey}', 'default_rate_limit_TPH', this.value ? parseInt(this.value) : null)" placeholder="Optional">
+                <small style="color: #a0a0a0; font-size: 12px; display: block; margin-top: 5px;">Tokens per hour limit</small>
+            </div>
+            
+            <div class="form-group">
+                <label>Default Rate Limit TPD</label>
+                <input type="number" value="${autoselect.default_rate_limit_TPD || ''}" onchange="updateAutoselect('${autoselectKey}', 'default_rate_limit_TPD', this.value ? parseInt(this.value) : null)" placeholder="Optional">
+                <small style="color: #a0a0a0; font-size: 12px; display: block; margin-top: 5px;">Tokens per day limit</small>
+            </div>
+            
+            <div class="form-group">
+                <label>Default Condense Context</label>
+                <input type="number" value="${autoselect.default_condense_context || ''}" onchange="updateAutoselect('${autoselectKey}', 'default_condense_context', this.value ? parseInt(this.value) : null)" placeholder="Optional">
+                <small style="color: #a0a0a0; font-size: 12px; display: block; margin-top: 5px;">Trigger context condensation at this token count</small>
+            </div>
+        </div>
+        
         <h4 style="margin-top: 20px; margin-bottom: 10px;">Available Models</h4>
         <p style="color: #a0a0a0; font-size: 14px; margin-bottom: 10px;">Define which models can be selected and their descriptions for AI analysis</p>
         <div id="models-${autoselectKey}"></div>

templates/dashboard/user_index.html

@@ -15,6 +15,83 @@
     <div class="alert alert-error">{{ error }}</div>
     {% endif %}
 
+    <!-- API Documentation Section -->
+    <div class="card">
+        <h2>🔌 Your API Endpoints</h2>
+        <p>Use your API token to access your personal configurations. Include the token in the Authorization header:</p>
+        <pre style="background: #1a1a2e; padding: 1rem; border-radius: 4px; overflow-x: auto;"><code>Authorization: Bearer YOUR_API_TOKEN</code></pre>
+        
+        <div class="api-endpoints">
+            <h3>Your Models</h3>
+            <div class="endpoint">
+                <code class="method GET">GET</code>
+                <code class="url">/api/user/models</code>
+                <p>List all models from your providers, rotations, and autoselects</p>
+            </div>
+            
+            <h3>Your Providers</h3>
+            <div class="endpoint">
+                <code class="method GET">GET</code>
+                <code class="url">/api/user/providers</code>
+                <p>List all your configured providers</p>
+            </div>
+            
+            <h3>Your Rotations</h3>
+            <div class="endpoint">
+                <code class="method GET">GET</code>
+                <code class="url">/api/user/rotations</code>
+                <p>List all your configured rotations</p>
+            </div>
+            
+            <h3>Your Autoselects</h3>
+            <div class="endpoint">
+                <code class="method GET">GET</code>
+                <code class="url">/api/user/autoselects</code>
+                <p>List all your configured autoselects</p>
+            </div>
+            
+            <h3>Your Chat Completions</h3>
+            <div class="endpoint">
+                <code class="method POST">POST</code>
+                <code class="url">/api/user/chat/completions</code>
+                <p>Send chat requests using your configurations</p>
+                <p class="example">Example model formats:</p>
+                <ul>
+                    <li><code>user-provider/myprovider/mymodel</code></li>
+                    <li><code>user-rotation/myrotation</code></li>
+                    <li><code>user-autoselect/myautoselect</code></li>
+                </ul>
+            </div>
+            
+            <h3>MCP Tools</h3>
+            <div class="endpoint">
+                <code class="method GET">GET</code>
+                <code class="url">/mcp/tools</code>
+                <p>List available MCP tools for your configurations</p>
+            </div>
+            <div class="endpoint">
+                <code class="method POST">POST</code>
+                <code class="url">/mcp/tools/call</code>
+                <p>Call MCP tools to manage your configurations</p>
+            </div>
+        </div>
+        
+        {% if session.role == 'admin' %}
+        <div class="admin-notice">
+            <h4>⚡ Admin Access</h4>
+            <p>As an admin user, you also have access to global configurations (providers, rotations, autoselects configured by the admin) in addition to your own user configurations.</p>
+            <p class="example">Admin model formats:</p>
+            <ul>
+                <li><code>provider/model</code> - global provider</li>
+                <li><code>rotation/myrotation</code> - global rotation</li>
+                <li><code>autoselect/myautoselect</code> - global autoselect</li>
+            </ul>
+        </div>
+        {% endif %}
+        
+        <p class="note">⚠️ Note: Your API token is required for all these endpoints. Manage your tokens in the <a href="{{ url_for(request, '/dashboard/user/tokens') }}">API Tokens</a> section.</p>
+    </div>
+
     <!-- Usage Statistics -->
     <div class="card">
         <h2>Usage Statistics</h2>
@@ -136,5 +213,116 @@
 .table tbody tr:hover {
     background: #1a1a2e;
 }
+
+/* API Endpoints Section */
+.api-endpoints {
+    margin-top: 1rem;
+}
+
+.api-endpoints h3 {
+    margin-top: 1.5rem;
+    color: #e0e0e0;
+    font-size: 1.1rem;
+}
+
+.endpoint {
+    background: #0f3460;
+    padding: 1rem;
+    margin: 0.5rem 0;
+    border-radius: 4px;
+    border-left: 3px solid #667eea;
+}
+
+.endpoint .method {
+    display: inline-block;
+    padding: 0.25rem 0.5rem;
+    border-radius: 3px;
+    font-weight: bold;
+    font-size: 0.8rem;
+    margin-right: 0.5rem;
+}
+
+.endpoint .method.GET {
+    background: #28a745;
+    color: white;
+}
+
+.endpoint .method.POST {
+    background: #007bff;
+    color: white;
+}
+
+.endpoint .url {
+    color: #e0e0e0;
+    font-family: monospace;
+}
+
+.endpoint p {
+    margin: 0.5rem 0 0 0;
+    color: #a0a0a0;
+    font-size: 0.9rem;
+}
+
+.endpoint .example {
+    color: #667eea;
+}
+
+.endpoint ul {
+    margin: 0.5rem 0 0 0;
+    padding-left: 1.5rem;
+    color: #a0a0a0;
+}
+
+.endpoint code {
+    background: #1a1a2e;
+    padding: 0.2rem 0.4rem;
+    border-radius: 3px;
+    font-size: 0.9rem;
+}
+
+.note {
+    margin-top: 1rem;
+    padding: 1rem;
+    background: #1a1a2e;
+    border-radius: 4px;
+    color: #a0a0a0;
+}
+
+.note a {
+    color: #667eea;
+}
+
+.admin-notice {
+    margin-top: 1.5rem;
+    padding: 1rem;
+    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+    border-radius: 8px;
+    color: white;
+}
+
+.admin-notice h4 {
+    margin: 0 0 0.5rem 0;
+    color: white;
+}
+
+.admin-notice p {
+    margin: 0.5rem 0;
+    color: white;
+}
+
+.admin-notice .example {
+    color: #ffd700;
+}
+
+.admin-notice ul {
+    margin: 0.5rem 0 0 0;
+    padding-left: 1.5rem;
+}
+
+.admin-notice code {
+    background: rgba(0,0,0,0.3);
+    padding: 0.2rem 0.4rem;
+    border-radius: 3px;
+}
 </style>
 {% endblock %}

templates/dashboard/user_tokens.html

@@ -15,6 +15,74 @@
     <div class="alert alert-error">{{ error }}</div>
     {% endif %}
 
+    <!-- API Documentation -->
+    <div class="card">
+        <h2>🔌 How to Use Your Token</h2>
+        <p>Include your API token in the <code>Authorization</code> header when making requests:</p>
+        <pre style="background: #1a1a2e; padding: 1rem; border-radius: 4px; overflow-x: auto;"><code>Authorization: Bearer YOUR_API_TOKEN</code></pre>
+        
+        <h3>Available Endpoints:</h3>
+        <table class="endpoints-table">
+            <thead>
+                <tr>
+                    <th>Method</th>
+                    <th>Endpoint</th>
+                    <th>Description</th>
+                </tr>
+            </thead>
+            <tbody>
+                <tr>
+                    <td><code class="method GET">GET</code></td>
+                    <td><code>/api/user/models</code></td>
+                    <td>List your models</td>
+                </tr>
+                <tr>
+                    <td><code class="method GET">GET</code></td>
+                    <td><code>/api/user/providers</code></td>
+                    <td>List your providers</td>
+                </tr>
+                <tr>
+                    <td><code class="method GET">GET</code></td>
+                    <td><code>/api/user/rotations</code></td>
+                    <td>List your rotations</td>
+                </tr>
+                <tr>
+                    <td><code class="method GET">GET</code></td>
+                    <td><code>/api/user/autoselects</code></td>
+                    <td>List your autoselects</td>
+                </tr>
+                <tr>
+                    <td><code class="method POST">POST</code></td>
+                    <td><code>/api/user/chat/completions</code></td>
+                    <td>Chat using your configs</td>
+                </tr>
+                <tr>
+                    <td><code class="method GET">GET</code></td>
+                    <td><code>/mcp/tools</code></td>
+                    <td>List MCP tools</td>
+                </tr>
+                <tr>
+                    <td><code class="method POST">POST</code></td>
+                    <td><code>/mcp/tools/call</code></td>
+                    <td>Call MCP tools</td>
+                </tr>
+            </tbody>
+        </table>
+        
+        <h3>Example curl commands:</h3>
+        <pre style="background: #1a1a2e; padding: 1rem; border-radius: 4px; overflow-x: auto;"><code># List your models
+curl -H "Authorization: Bearer YOUR_TOKEN" http://localhost:17765/api/user/models
+
+# List your providers  
+curl -H "Authorization: Bearer YOUR_TOKEN" http://localhost:17765/api/user/providers
+
+# Send a chat request
+curl -X POST -H "Authorization: Bearer YOUR_TOKEN" \
+  -H "Content-Type: application/json" \
+  -d '{"model": "user-rotation/myrotation", "messages": [{"role": "user", "content": "Hello"}]}' \
+  http://localhost:17765/api/user/chat/completions</code></pre>
+    </div>
+
     <div class="card">
         <h2>API Tokens</h2>
         <div id="tokens-list">
@@ -356,5 +424,42 @@ renderTokens();
     font-family: monospace;
     word-break: break-all;
 }
+
+/* API Endpoints Table */
+.endpoints-table {
+    width: 100%;
+    margin: 1rem 0;
+    border-collapse: collapse;
+}
+
+.endpoints-table th,
+.endpoints-table td {
+    padding: 0.75rem;
+    text-align: left;
+    border-bottom: 1px solid #333;
+}
+
+.endpoints-table th {
+    background: #1a1a2e;
+    font-weight: bold;
+}
+
+.endpoints-table .method {
+    display: inline-block;
+    padding: 0.2rem 0.5rem;
+    border-radius: 3px;
+    font-weight: bold;
+    font-size: 0.75rem;
+}
+
+.endpoints-table .method.GET {
+    background: #28a745;
+    color: white;
+}
+
+.endpoints-table .method.POST {
+    background: #007bff;
+    color: white;
+}
 </style>
 {% endblock %}
\ No newline at end of file