perf(agent): cap recall + 2s deadline so LLM fallback always fires

Trace showed every Assistant submit triggered 60+ sequential
GET /api/ea2/flow/{key} requests via agentMemory.recall before
llmClient.chat ran. With a slow EA2 demo, recall took longer
than the test wait so the LLM fetch never started, the agent
reply never rendered, and [llm] telemetry never logged.

Two fixes:

1. agentMemory.listAll(cap) - default 200, recall uses 30.
   Edges are returned newest-first by EA2 so the 30-cap keeps the
   most recent vault entries and drops the long-tail history that
   was pinning the network.

2. routeAgentInput wraps recall in a 2s AbortController. If recall
   doesn't finish in 2s, hints[] stays empty and llmFallback runs
   immediately. Recall is best-effort context, never a blocker.

src/lib/agentMemory.ts

@@ -97,7 +97,7 @@ export const agentMemory = {
     return mem._key;
   },
 
-  async listAll(email: string, signal?: AbortSignal): Promise<Memory[]> {
+  async listAll(email: string, signal?: AbortSignal, cap = 200): Promise<Memory[]> {
     const vaultKey = vaultKeyFor(email);
     const ok = await ensureVault(vaultKey, email, signal);
     if (!ok) return [];
@@ -107,7 +107,8 @@ export const agentMemory = {
     const memKeys = (edges as any[])
       .filter((e) => e?.role === "presentation")
       .map((e) => (e._to || "").split("/").pop())
-      .filter(Boolean) as string[];
+      .filter(Boolean)
+      .slice(0, cap) as string[];
     const memos = await Promise.all(
       memKeys.map(async (k) => {
         try {
@@ -132,7 +133,7 @@ export const agentMemory = {
    * Score = count of unique stem matches. Ties broken by recency.
    */
   async recall(email: string, query: string, limit = 5, signal?: AbortSignal): Promise<Memory[]> {
-    const all = await agentMemory.listAll(email, signal);
+    const all = await agentMemory.listAll(email, signal, 30);
     if (!query.trim() || all.length === 0) return all.slice(0, limit);
     const stems = new Set(
       query.toLowerCase().replace(/[^a-z0-9\s]/g, " ").split(/\s+/).filter((w) => w.length > 2)