feat(backend): integrate PostHog analytics for enhanced observability

- Added PostHog configuration options to .env.example files for both Docker and Surfsense backend.
- Introduced PostHog dependency in pyproject.toml.
- Implemented analytics middleware to capture various events across the application, including user authentication, automation runs, and API requests.
- Enhanced existing routes and services to emit analytics events, providing insights into user interactions and system performance.
- Ensured graceful shutdown of analytics clients in worker processes and application lifecycles.

surfsense_mcp/mcp_server/core/client.py

@@ -37,7 +37,11 @@ class SurfSenseClient:
         self._fallback_api_key = fallback_api_key
         self._http = httpx.AsyncClient(
             base_url=api_base,
-            headers={"Accept": "application/json"},
+            # ``X-SurfSense-Client`` lets the backend distinguish PAT traffic
+            # originating from this MCP server vs. raw PAT scripts, so
+            # "documents added via MCP" / "searches via MCP" are queryable.
+            # Server-to-server, so no CORS implications.
+            headers={"Accept": "application/json", "X-SurfSense-Client": "mcp"},
             timeout=timeout,
         )