feat(ai): local LLM auto-discovery — find a running Ollama, suggest connecting

New lib/ai/local-discovery.ts: one /api/tags query against the loopback
addresses Ollama binds to (127.0.0.1/localhost), no follow-up /api/show
round trips needed — the tags response already carries capabilities, size,
and parameter_size, enough to recommend a default model. Picks the
smallest non-"thinking" chat-capable model for the fastest first response
("Connect" pre-fills provider+baseUrl+model in one click), and separately
surfaces the largest as a "higher quality" alternative.

New banner in ai-assistant-settings.tsx: fires when Local isn't yet
configured, offers one-click Connect or a persisted "Not now" dismissal.

13 new unit tests using this machine's actual Ollama /api/tags response
(11 real installed models — qwen2.5:32b, llama3.2, deepseek-r1 x2,
gemma4 x3, hermes3, qwen3, qwen3.5, nomic-embed-text) as literal fixtures,
per the explicit instruction to use this machine as the test case:
confirms exactly one query is required, the heuristic recommends
llama3.2:latest (fastest) / qwen2.5:32b (largest) on this real fleet,
never recommends an embedding-only model, and degrades correctly when a
candidate base URL is unreachable.

Full QA gate: tsc clean, eslint clean, 2498/2498 tests passing, build clean.
Also live-verified in a real browser session against this machine's real
Ollama — the banner rendered with exactly these two model names.

components/settings/ai-assistant-settings.tsx

@@ -1,7 +1,7 @@
 'use client';
 
 import { useCallback, useEffect, useMemo, useState } from 'react';
-import { RefreshCw, CheckCircle, AlertTriangle, Loader2, Plus, Trash2 } from 'lucide-react';
+import { RefreshCw, CheckCircle, AlertTriangle, Loader2, Plus, Trash2, Sparkles, X } from 'lucide-react';
 import { SettingsSection, SettingItem, ToggleSwitch, RadioGroup, Select } from './settings-section';
 import { Button } from '@/components/ui/button';
 import { apiFetch } from '@/lib/browser-navigation';
@@ -9,6 +9,14 @@ import { DEFAULT_AI_POLICY, type AiPolicy } from '@/lib/ai/types';
 import { supportsLocalLlm, localLlmNeedsCorsSetup } from '@/lib/platform-capabilities';
 import { getAiApiKey, setAiApiKey, clearAiApiKey } from '@/lib/ai/key-store';
 import { loadAiSettings, saveAiSettings, createProfile, type AiLocalSettings } from '@/lib/ai/local-settings';
+import {
+  discoverLocalOllama,
+  recommendDefaultModel,
+  largestModel,
+  isLocalDiscoveryDismissed,
+  dismissLocalDiscovery,
+  type LocalDiscoveryResult,
+} from '@/lib/ai/local-discovery';
 import {
   askMail,
   listLocalModels,
@@ -91,6 +99,47 @@ export function AiAssistantSettings() {
     }
   }, [settings.localBaseUrl]);
 
+  // ── Local discovery — proactively find an already-running Ollama and
+  // offer a one-click connect, rather than making the user hunt down and
+  // type a base URL + model name by hand. ──
+  const [discovery, setDiscovery] = useState<LocalDiscoveryResult | null>(null);
+  const [discoveryDismissed, setDiscoveryDismissed] = useState(true);
+
+  useEffect(() => {
+    setDiscoveryDismissed(isLocalDiscoveryDismissed());
+  }, []);
+
+  useEffect(() => {
+    if (!canUseLocal || discoveryDismissed || settings.localModel) return;
+    let cancelled = false;
+    (async () => {
+      const result = await discoverLocalOllama();
+      if (!cancelled) setDiscovery(result);
+    })();
+    return () => {
+      cancelled = true;
+    };
+  }, [canUseLocal, discoveryDismissed, settings.localModel]);
+
+  const connectDiscoveredLocal = useCallback(() => {
+    if (!discovery) return;
+    const recommended = recommendDefaultModel(discovery.models) ?? discovery.models[0]?.name ?? null;
+    if (!recommended) return;
+    setSettings((prev) => {
+      const next: AiLocalSettings = { ...prev, provider: 'local', localBaseUrl: discovery.baseUrl, localModel: recommended };
+      saveAiSettings(next);
+      return next;
+    });
+    setLocalModels(discovery.models.filter((m) => m.capabilities.includes('completion')).map((m) => m.name));
+    setDiscovery(null);
+  }, [discovery]);
+
+  const dismissDiscoveryBanner = useCallback(() => {
+    dismissLocalDiscovery();
+    setDiscoveryDismissed(true);
+    setDiscovery(null);
+  }, []);
+
   // ── Server provider ──
   const [serverModels, setServerModels] = useState<string[]>([]);
   const [refreshingServer, setRefreshingServer] = useState(false);
@@ -209,8 +258,35 @@ export function AiAssistantSettings() {
     );
   }
 
+  const discoveryRecommended = discovery ? recommendDefaultModel(discovery.models) : null;
+  const discoveryLargest = discovery ? largestModel(discovery.models) : null;
+
   return (
     <div className="space-y-6">
+      {discovery && discoveryRecommended && (
+        <div className="flex items-start gap-3 rounded-lg border border-primary/30 bg-primary/5 p-4">
+          <Sparkles className="w-5 h-5 mt-0.5 text-primary shrink-0" />
+          <div className="flex-1 min-w-0 space-y-2">
+            <p className="text-sm font-medium text-foreground">Local AI found on this machine</p>
+            <p className="text-sm text-muted-foreground">
+              Ollama is running at {discovery.baseUrl} with {discovery.models.length} model{discovery.models.length === 1 ? '' : 's'} installed.
+              Recommended for quick answers: <span className="font-medium text-foreground">{discoveryRecommended}</span>.
+              {discoveryLargest && discoveryLargest !== discoveryRecommended && (
+                <> Also available for higher-quality answers: <span className="font-medium text-foreground">{discoveryLargest}</span>.</>
+              )}
+            </p>
+            <div className="flex gap-2">
+              <Button size="sm" onClick={connectDiscoveredLocal}>
+                <Sparkles className="w-3.5 h-3.5 me-1.5" /> Connect
+              </Button>
+              <Button size="sm" variant="outline" onClick={dismissDiscoveryBanner}>
+                <X className="w-3.5 h-3.5 me-1.5" /> Not now
+              </Button>
+            </div>
+          </div>
+        </div>
+      )}
+
       <SettingsSection
         title="AI Assistant"
         description="Ask questions about your synced mail. Local runs entirely on this machine's own model runtime; server is centrally hosted and licensed per seat; public sends your question to a provider you choose, using your own API key."