Merge pull request 'perf(citation-verification): cache view + parallelize backend retrieval' (#369) from worktree-citation-verify-perf into main

mcp-server/src/legal_mcp/services/case_citation_verification.py

@@ -18,6 +18,7 @@ rows. Reuses the one corpus search + the one authority query + the one radar (G2
 
 from __future__ import annotations
 
+import asyncio
 import logging
 from uuid import UUID
 
@@ -65,25 +66,37 @@ async def build_view(case_number: str) -> dict:
     except Exception as e:  # noqa: BLE001 — radar is best-effort
         logger.warning("citation_verification radar failed for %s: %s", case_number, e)
 
-    out_args: list[dict] = []
+    async def _fetch(a: dict) -> tuple[list[dict], dict]:
-    n_verified = 0
+        """Per-argument corpus search + batched authority — run concurrently.
-    for a in args:
+
-        aid = str(a["id"])
+        Each call is one Voyage embed + one vector search (+ one batched authority
+        query); independent across arguments, so they fan out rather than waterfall.
+        """
         title = (a.get("argument_title") or "").strip()
         topic = (a.get("legal_topic") or "").strip()
         query = f"{ctx} {title}. {topic}".strip()
-
+        hits: list[dict] = []
-        hits = []
         try:
             hits = await precedent_library.search_library(
                 query=query, limit=_SUGGEST_PER_ISSUE, include_halachot=True)
         except Exception as e:  # noqa: BLE001
             logger.warning("citation_verification search failed (%s): %s", title[:30], e)
-
         # Resolve the authority breakdown for the hit set in one batched query.
         clids = [UUID(str(h["case_law_id"])) for h in hits
                  if h.get("case_law_id") and float(h.get("score", 0) or 0) >= _SUGGEST_FLOOR]
         authority = await db.citation_authority(clids) if clids else {}
+        return hits, authority
+
+    # Fan out the expensive per-argument retrieval concurrently — was a sequential
+    # waterfall (N args × Voyage embed + vector search each). gather preserves order.
+    fetched = await asyncio.gather(*(_fetch(a) for a in args)) if args else []
+
+    out_args: list[dict] = []
+    n_verified = 0
+    for a, (hits, authority) in zip(args, fetched):
+        aid = str(a["id"])
+        title = (a.get("argument_title") or "").strip()
+        topic = (a.get("legal_topic") or "").strip()
 
         seen: set[str] = set()
         supporting: list[dict] = []

web-ui/src/lib/api/citation-verification.ts

@@ -84,7 +84,13 @@ export function useCitationVerification(caseNumber: string | undefined) {
         { signal },
       ),
     enabled: Boolean(caseNumber),
-    staleTime: 10_000,
+    // The view is an expensive synthesized build (per-argument Voyage embed +
+    // vector search on the backend). Keep it fresh for 5 min and do NOT refetch
+    // on remount — Radix Tabs unmounts inactive TabsContent, so every tab switch
+    // back used to trigger a full rebuild. Verify mutations invalidate the query,
+    // so edits still pull fresh data.
+    staleTime: 5 * 60_000,
+    refetchOnMount: false,
   });
 }