mcp-souschef 2.2.0__py3-none-any.whl → 2.8.0__py3-none-any.whl

souschef/ui/health_check.py

@@ -0,0 +1,36 @@
+#!/usr/bin/env python3
+"""Health check endpoint for SousChef UI Docker container."""
+
+import json
+import sys
+from pathlib import Path
+
+# Add the app directory to the path for proper imports
+app_path = Path(__file__).parent.parent
+if str(app_path) not in sys.path:
+    sys.path.insert(0, str(app_path))
+
+
+def main():
+    """Run health check."""
+    try:
+        # Try to import a core module to verify the environment is working
+        from souschef.core.constants import VERSION
+
+        sys.stdout.write(
+            json.dumps(
+                {"status": "healthy", "service": "souschef-ui", "version": VERSION}
+            )
+        )
+        sys.exit(0)
+    except Exception as e:
+        sys.stdout.write(
+            json.dumps(
+                {"status": "unhealthy", "service": "souschef-ui", "error": str(e)}
+            )
+        )
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    main()

souschef/ui/pages/ai_settings.py

@@ -0,0 +1,497 @@
+"""
+AI Settings Page for SousChef UI.
+
+Configure and validate AI provider settings for the SousChef MCP server.
+"""
+
+import json
+from pathlib import Path
+from urllib.parse import urlparse, urlunparse
+
+import streamlit as st
+
+# AI Provider Constants
+ANTHROPIC_PROVIDER = "Anthropic (Claude)"
+OPENAI_PROVIDER = "OpenAI (GPT)"
+WATSON_PROVIDER = "IBM Watsonx"
+LIGHTSPEED_PROVIDER = "Red Hat Lightspeed"
+LOCAL_PROVIDER = "Local Model"
+
+# UI Constants
+API_KEY_LABEL = "API Key"
+
+# Import AI libraries (optional dependencies)
+try:
+    import anthropic
+except ImportError:
+    anthropic = None  # type: ignore[assignment]
+
+try:
+    from ibm_watsonx_ai import APIClient  # type: ignore[import-not-found]
+except ImportError:
+    APIClient = None
+
+try:
+    import requests  # type: ignore[import-untyped]
+except ImportError:
+    requests = None
+
+try:
+    import openai
+except ImportError:
+    openai = None  # type: ignore[assignment]
+
+
+def _get_model_options(provider):
+    """Get model options for the selected provider."""
+    if provider == ANTHROPIC_PROVIDER:
+        return [
+            "claude-3-5-sonnet-20241022",
+            "claude-3-5-haiku-20241022",
+            "claude-3-opus-20240229",
+        ]
+    elif provider == OPENAI_PROVIDER:
+        return ["gpt-4o", "gpt-4o-mini", "gpt-4-turbo", "gpt-3.5-turbo"]
+    elif provider == WATSON_PROVIDER:
+        return [
+            "meta-llama/llama-3-70b-instruct",
+            "meta-llama/llama-3-8b-instruct",
+            "ibm/granite-13b-instruct-v2",
+            "ibm/granite-13b-chat-v2",
+        ]
+    elif provider == LIGHTSPEED_PROVIDER:
+        return ["codellama/CodeLlama-34b-Instruct-hf"]
+    else:
+        return ["local-model"]
+
+
+def _render_api_configuration(provider):
+    """Render API configuration UI and return config values."""
+    if provider == LOCAL_PROVIDER:
+        st.info("Local model configuration will be added in a future update.")
+        return "", "", ""
+    elif provider == WATSON_PROVIDER:
+        col1, col2, col3 = st.columns(3)
+        with col1:
+            api_key = st.text_input(
+                API_KEY_LABEL,
+                type="password",
+                help="Enter your IBM Watsonx API key",
+                key="api_key_input",
+                placeholder="your-watsonx-api-key",
+            )
+        with col2:
+            project_id = st.text_input(
+                "Project ID",
+                type="password",
+                help="Enter your IBM Watsonx Project ID",
+                key="project_id_input",
+                placeholder="your-project-id",
+            )
+        with col3:
+            base_url = st.text_input(
+                "Base URL",
+                help="IBM Watsonx API base URL",
+                key="base_url_input",
+                placeholder="https://us-south.ml.cloud.ibm.com",
+            )
+        return api_key, base_url, project_id
+    elif provider == LIGHTSPEED_PROVIDER:
+        col1, col2 = st.columns(2)
+        with col1:
+            api_key = st.text_input(
+                API_KEY_LABEL,
+                type="password",
+                help="Enter your Red Hat Lightspeed API key",
+                key="api_key_input",
+                placeholder="your-lightspeed-api-key",
+            )
+        with col2:
+            base_url = st.text_input(
+                "Base URL",
+                help="Red Hat Lightspeed API base URL",
+                key="base_url_input",
+                placeholder="https://api.redhat.com",
+            )
+        return api_key, base_url, ""
+    else:
+        col1, col2 = st.columns(2)
+        with col1:
+            api_key = st.text_input(
+                API_KEY_LABEL,
+                type="password",
+                help=f"Enter your {provider.split(' ')[0]} API key",
+                key="api_key_input",
+                placeholder=f"sk-... (for {provider.split(' ')[0]})",
+            )
+        with col2:
+            if provider == OPENAI_PROVIDER:
+                base_url = st.text_input(
+                    "Base URL (Optional)",
+                    help="Custom OpenAI API base URL",
+                    key="base_url_input",
+                    placeholder="https://api.openai.com/v1",
+                )
+            else:
+                base_url = ""
+        return api_key, base_url, ""
+
+
+def _render_advanced_settings():
+    """Render advanced settings UI and return values."""
+    with st.expander("Advanced Settings"):
+        col1, col2 = st.columns(2)
+        with col1:
+            temperature = st.slider(
+                "Temperature",
+                min_value=0.0,
+                max_value=2.0,
+                value=0.7,
+                step=0.1,
+                help="Controls randomness in AI responses "
+                "(0.0 = deterministic, 2.0 = very random)",
+                key="temperature_slider",
+            )
+        with col2:
+            max_tokens = st.number_input(
+                "Max Tokens",
+                min_value=100,
+                max_value=100000,
+                value=4000,
+                help="Maximum number of tokens to generate",
+                key="max_tokens_input",
+            )
+    return temperature, max_tokens
+
+
+def _render_validation_section(
+    provider, api_key, model, base_url, project_id, temperature, max_tokens
+):
+    """Render validation and save buttons."""
+    st.subheader("Configuration Validation")
+    col1, col2 = st.columns([1, 1])
+    with col1:
+        if st.button("Validate Configuration", type="primary", width="stretch"):
+            validate_ai_configuration(provider, api_key, model, base_url, project_id)
+    with col2:
+        if st.button("Save Settings", width="stretch"):
+            save_ai_settings(
+                provider, api_key, model, base_url, temperature, max_tokens, project_id
+            )
+
+
+def show_ai_settings_page():
+    """Show the AI settings configuration page."""
+    st.markdown("""
+    Configure your AI provider settings for the SousChef MCP server.
+    These settings determine which AI model will be used for Chef to Ansible
+    conversions.
+    """)
+
+    # AI Provider Selection
+    st.subheader("AI Provider Configuration")
+
+    col1, col2 = st.columns([1, 2])
+
+    with col1:
+        ai_provider = st.selectbox(
+            "AI Provider",
+            [
+                ANTHROPIC_PROVIDER,
+                OPENAI_PROVIDER,
+                WATSON_PROVIDER,
+                LIGHTSPEED_PROVIDER,
+                LOCAL_PROVIDER,
+            ],
+            help="Select your preferred AI provider",
+            key="ai_provider_select",
+        )
+
+    with col2:
+        model_options = _get_model_options(ai_provider)
+        selected_model = st.selectbox(
+            "Model",
+            model_options,
+            help="Select the AI model to use",
+            key="ai_model_select",
+        )
+
+    # API Configuration
+    st.subheader("API Configuration")
+    api_key, base_url, project_id = _render_api_configuration(ai_provider)
+
+    # Advanced Settings
+    temperature, max_tokens = _render_advanced_settings()
+
+    # Validation Section
+    _render_validation_section(
+        ai_provider,
+        api_key,
+        selected_model,
+        base_url,
+        project_id,
+        temperature,
+        max_tokens,
+    )
+
+    # Current Settings Display
+    display_current_settings()
+
+
+def validate_ai_configuration(provider, api_key, model, base_url="", project_id=""):
+    """Validate the AI configuration by making a test API call."""
+    if not api_key and provider != "Local Model":
+        st.error("API key is required for validation.")
+        return
+
+    if provider == WATSON_PROVIDER and not project_id:
+        st.error("Project ID is required for IBM Watsonx validation.")
+        return
+
+    with st.spinner("Validating AI configuration..."):
+        try:
+            if provider == ANTHROPIC_PROVIDER:
+                success, message = validate_anthropic_config(api_key, model)
+            elif provider == OPENAI_PROVIDER:
+                success, message = validate_openai_config(api_key, model, base_url)
+            elif provider == WATSON_PROVIDER:
+                success, message = validate_watson_config(api_key, project_id, base_url)
+            elif provider == LIGHTSPEED_PROVIDER:
+                success, message = validate_lightspeed_config(api_key, model, base_url)
+            else:
+                st.info("Local model validation not implemented yet.")
+                return
+
+            if success:
+                st.success(f"Configuration validated successfully! {message}")
+            else:
+                st.error(f"Validation failed: {message}")
+
+        except Exception as e:
+            st.error(f"Validation error: {str(e)}")
+
+
+def _sanitize_lightspeed_base_url(base_url: str) -> str:
+    """
+    Sanitize and validate the Red Hat Lightspeed base URL to prevent SSRF.
+
+    - Default to the standard Lightspeed endpoint if no URL is provided.
+    - Only allow HTTPS scheme.
+    - Restrict host to known Red Hat-owned Lightspeed domains.
+    - Strip any user-supplied path, query, or fragment.
+    """
+    default_url = "https://api.redhat.com"
+    allowed_hosts = {
+        "api.redhat.com",
+    }
+
+    if not base_url or not str(base_url).strip():
+        return default_url
+
+    parsed = urlparse(base_url)
+
+    # If scheme is missing, assume https
+    if not parsed.scheme:
+        parsed = parsed._replace(scheme="https")
+
+    if parsed.scheme.lower() != "https":
+        raise ValueError("Base URL must use HTTPS.")
+
+    hostname = (parsed.hostname or "").lower()
+    if hostname not in allowed_hosts:
+        raise ValueError("Base URL host must be a supported Red Hat domain.")
+
+    # Normalize to scheme + netloc only; drop path/query/fragment.
+    cleaned = parsed._replace(path="", params="", query="", fragment="")
+    return urlunparse(cleaned)
+
+
+def validate_anthropic_config(api_key, model):
+    """Validate Anthropic API configuration."""
+    if anthropic is None:
+        return False, "Anthropic library not installed"
+
+    try:
+        client = anthropic.Anthropic(api_key=api_key)
+
+        # Make a simple test call
+        client.messages.create(
+            model=model, max_tokens=10, messages=[{"role": "user", "content": "Hello"}]
+        )
+
+        return True, f"Successfully connected to {model}"
+
+    except Exception as e:
+        return False, f"Connection failed: {e}"
+
+
+def validate_openai_config(api_key, model, base_url=""):
+    """Validate OpenAI API configuration."""
+    if openai is None:
+        return False, "OpenAI library not installed. Run: pip install openai"
+
+    try:
+        client_kwargs = {"api_key": api_key}
+        if base_url:
+            client_kwargs["base_url"] = base_url
+
+        client = openai.OpenAI(**client_kwargs)
+
+        # Make a simple test call
+        client.chat.completions.create(
+            model=model, messages=[{"role": "user", "content": "Hello"}], max_tokens=5
+        )
+
+        return True, f"Successfully connected to {model}"
+
+    except Exception as e:
+        return False, f"Connection failed: {e}"
+
+
+def validate_lightspeed_config(api_key, model, base_url=""):
+    """Validate Red Hat Lightspeed API configuration."""
+    if requests is None:
+        return False, "Requests library not installed. Run: pip install requests"
+
+    try:
+        # Sanitize and validate base URL
+        sanitized_url = _sanitize_lightspeed_base_url(base_url)
+
+        # Make a simple test request to validate API key
+        headers = {
+            "Authorization": f"Bearer {api_key}",
+            "Content-Type": "application/json",
+        }
+
+        # Test with a simple completion request
+        test_payload = {
+            "model": model,
+            "prompt": "Hello",
+            "max_tokens": 5,
+        }
+
+        response = requests.post(
+            f"{sanitized_url}/v1/completions",
+            headers=headers,
+            json=test_payload,
+            timeout=10,
+        )
+
+        if response.status_code == 200:
+            return True, f"Successfully connected to Red Hat Lightspeed {model}"
+        else:
+            return False, (
+                f"API request failed with status {response.status_code}: "
+                f"{response.text}"
+            )
+
+    except Exception as e:
+        return False, f"Connection failed: {e}"
+
+
+def validate_watson_config(api_key, project_id, base_url=""):
+    """Validate IBM Watsonx API configuration."""
+    if APIClient is None:
+        return False, (
+            "IBM Watsonx AI library not installed. Run: pip install ibm-watsonx-ai"
+        )
+
+    try:
+        # Initialize Watsonx client
+        client = APIClient(
+            api_key=api_key,
+            project_id=project_id,
+            url=base_url or "https://us-south.ml.cloud.ibm.com",
+        )
+
+        # Test connection by listing available models
+        models = client.foundation_models.get_model_specs()
+        if models:
+            return True, (
+                f"Successfully connected to IBM Watsonx. "
+                f"Found {len(models)} available models."
+            )
+        else:
+            return False, "Connected to IBM Watsonx but no models available."
+
+    except Exception as e:
+        return False, f"Connection failed: {e}"
+
+
+def save_ai_settings(
+    provider, api_key, model, base_url, temperature, max_tokens, project_id=""
+):
+    """Save AI settings to configuration file."""
+    try:
+        # Use /tmp/.souschef for container compatibility (tmpfs is writable)
+        config_dir = Path("/tmp/.souschef")
+        config_dir.mkdir(exist_ok=True)
+        config_file = config_dir / "ai_config.json"
+
+        config = {
+            "provider": provider,
+            "model": model,
+            "api_key": api_key if api_key else None,
+            "base_url": base_url if base_url else None,
+            "project_id": project_id if project_id else None,
+            "temperature": temperature,
+            "max_tokens": max_tokens,
+            "last_updated": str(st.session_state.get("timestamp", "Unknown")),
+        }
+
+        with config_file.open("w") as f:
+            json.dump(config, f, indent=2)
+
+        # Store in session state for immediate use
+        st.session_state.ai_config = config
+
+        st.success("Settings saved successfully!")
+
+    except Exception as e:
+        st.error(f"Failed to save settings: {str(e)}")
+
+
+def display_current_settings():
+    """Display current AI settings."""
+    st.subheader("Current Configuration")
+
+    # Try to load from file first, then session state
+    config = load_ai_settings()
+
+    if config:
+        col1, col2 = st.columns(2)
+
+        with col1:
+            st.metric("Provider", config.get("provider", "Not configured"))
+            st.metric("Model", config.get("model", "Not configured"))
+
+        with col2:
+            st.metric("Temperature", config.get("temperature", "Not set"))
+            st.metric("Max Tokens", config.get("max_tokens", "Not set"))
+
+        if config.get("last_updated"):
+            st.caption(f"Last updated: {config['last_updated']}")
+
+        # Security note
+        if config.get("api_key"):
+            st.info("API key is configured and stored securely.")
+        else:
+            st.warning("No API key configured.")
+    else:
+        st.info("No AI configuration found. Please configure your settings above.")
+
+
+def load_ai_settings():
+    """Load AI settings from configuration file."""
+    try:
+        # Use /tmp/.souschef for container compatibility (tmpfs is writable)
+        config_file = Path("/tmp/.souschef/ai_config.json")
+        if config_file.exists():
+            with config_file.open() as f:
+                return json.load(f)
+    except Exception as e:
+        # Failed to load config from file; fall back to session state/defaults
+        st.warning(f"Unable to load saved AI settings: {e}")
+
+    # Fallback to session state or return empty dict
+    return st.session_state.get("ai_config", {})