feat: made chat fast

- Introduced lazy knowledge base retrieval mode, allowing the main agent to fetch KB content on demand via the `search_knowledge_base` tool, improving performance by skipping expensive pre-injection processes.
- Added cross-thread caching capability, enabling reuse of compiled graphs across different user chats, reducing latency for returning users.
- Updated middleware to support new lazy loading and caching features, ensuring efficient resource utilization and improved response times.
- Enhanced logging for performance tracking during knowledge retrieval and agent interactions.

surfsense_web/app/api/zero/query/route.ts

@@ -12,7 +12,7 @@ import { schema } from "@/zero/schema";
 // container and would make every authenticated Zero query fail with a 503.
 const backendURL = (
 	process.env.FASTAPI_BACKEND_INTERNAL_URL ||
-	BACKEND_URL ||
+	process.env.BACKEND_URL ||
 	"http://localhost:8000"
 ).replace(/\/$/, "");