Merge pull request #4 from sridhar-daxa/main

Add support for pebblo mcp-server api-key

Author: sridhar-daxa

atlassian_langgraph_app/README.md

@@ -98,21 +98,27 @@ The application follows a clean modular design:
 ## Configuration
 
 ### Environment Variables
-The application uses environment variables for configuration. Create a `.env` file in the project root with the following variables:
+ The application uses environment variables for configuration. Create a `.env` file in the project root with the following variables:
 
 ```bash
 # OpenAI API Configuration
 OPENAI_API_KEY=your-api-key
 
 # MCP Server Configuration
 MCP_SERVER_URL=your-server-url
+# Optional: API key for MCP server (sent as Authorization: Bearer <key>)
+MCP_SERVER_API_KEY=your-mcp-api-key
 ```
 
 ### Required Environment Variables
 
 1. **OPENAI_API_KEY**: Your OpenAI API key for accessing GPT models
 2. **MCP_SERVER_URL**: The URL of your Atlassian MCP server (e.g., `http://localhost:9003/mcp`)
 
+### Optional Environment Variables
+
+- **MCP_SERVER_API_KEY**: If set, the app will add `Authorization: Bearer <MCP_SERVER_API_KEY>` to requests to the MCP server.
+
 ### Security Notes
 - Never commit your `.env` file to version control
 - Use different API keys for development and production

atlassian_langgraph_app/main.py

@@ -14,27 +14,36 @@
 # Get environment variables
 OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
 MCP_SERVER_URL = os.getenv("MCP_SERVER_URL")
+MCP_SERVER_API_KEY = os.getenv("MCP_SERVER_API_KEY")
 
 # Validate required environment variables
 if not OPENAI_API_KEY:
     raise ValueError("OPENAI_API_KEY environment variable is required")
 if not MCP_SERVER_URL:
     raise ValueError("MCP_SERVER_URL environment variable is required")
+if not MCP_SERVER_API_KEY:
+    raise ValueError("MCP_SERVER_API_KEY environment variable is required")
 
 # Load chat model
 chat_model = ChatOpenAI(model="gpt-4o-mini")
 
 async def setup_langgraph():
     """Setup and return the LangGraph with Atlassian MCP tools"""
     # Connect to Atlassian MCP server via Docker
-    mcp_client = MultiServerMCPClient(
-        {
-            "mcp-atlassian": {
-                "url": MCP_SERVER_URL,
-                "transport": "streamable_http" 
-            }
+    server_config = {
+        "url": MCP_SERVER_URL,
+        "transport": "streamable_http",
+    }
+
+    # Optionally pass Authorization header if MCP_SERVER_API_KEY is set
+    if MCP_SERVER_API_KEY:
+        server_config["headers"] = {
+            "Authorization": f"Bearer {MCP_SERVER_API_KEY}"
         }
-    )
+
+    mcp_client = MultiServerMCPClient({
+        "mcp-atlassian": server_config
+    })
 
     # Get tools from the client
     tools = await mcp_client.get_tools()