refactor(ai): remove semantic embedding cache + Qdrant vector store

Hapus seluruh fitur embedding/Qdrant (tidak dipakai lagi):

- gateway: drop embeddingClient.ts, qdrantClient.ts, archiveEmbedder.ts
  dan tes qdrantEnsure.test.ts; moderationOrchestrator kembali ke
  exact-hash cache -> LLM (tanpa phase-2 semantic lookup); textCacheStore
  kehilangan findSimilarTextModeration / parseQdrantVerdict /
  isSemanticBandAccepted / upsertBareKeyToQdrant; cache-prune hanya
  menyapu Postgres.
- backend: drop embed.ts + qdrant.ts, endpoint messages.semanticSearch
  dan schema/type terkait; kolom embedding dilepas dari schema
  text_analysis_cache.
- frontend: hapus toggle EXACT/SEMANTIC, hook useSemanticSearch,
  API client + tipe SemanticSearchResult.
- config: buang AI_LLM_EMBEDDING_* dan QDRANT_* (env + .env.example).
- docs: ARCHITECTURE.md / AGENTS.md / README.md / diagram arsitektur
  disesuaikan (LLM caller - vision, cache = exact-hash saja).

Verifikasi: tsc 0 (backend, gateway, frontend); bun test 135 pass +
37 pass, 0 fail; biome 0 error.

services/frontend/src/app/(dashboard)/messages/view.tsx

@@ -11,7 +11,6 @@ import {
   Paperclip,
   Search,
   ShieldAlert,
-  Sparkles,
 } from "lucide-react";
 import { useCallback, useEffect, useMemo, useRef, useState } from "react";
 import { ActivityHeatmap } from "@/components/ActivityHeatmap";
@@ -42,7 +41,6 @@ import {
   useMessagesWsSync,
   useRecentEdits,
   useReviewWsSync,
-  useSemanticSearch,
 } from "@/hooks";
 import { useStaggerReveal } from "@/hooks/use-gsap-animation";
 import { aiTone } from "@/lib/ai-status";
@@ -85,9 +83,6 @@ export function MessagesView({
   const [channelId, setChannelId] = useState<string | null>(null);
   const [selected, setSelected] = useState<string | null>(null);
   const [query, setQuery] = useState("");
-  // Search mode: "exact" (substring match over captured messages) or
-  // "semantic" (vector similarity over the persistent Qdrant archive).
-  const [semanticMode, setSemanticMode] = useState(false);
   // feed | timeline: "timeline" groups messages into date-grouped cards.
   const [viewMode, setViewMode] = useState<"feed" | "timeline">("feed");
   // Guard against loading the entire history on a long scroll: cap how many
@@ -123,15 +118,7 @@ export function MessagesView({
   const loadMore = useLoadMore();
   useMessagesWsSync(ws, guildId ?? "");
   useReviewWsSync(ws);
-  const search = useMessageSearch(
-    query,
-    query.trim().length >= 2 && !semanticMode,
-  );
-  const semantic = useSemanticSearch(
-    query,
-    query.trim().length >= 2 && semanticMode,
-    guildId,
-  );
+  const search = useMessageSearch(query, query.trim().length >= 2);
   const activity = useMessageActivity(30);
   const edits = useRecentEdits(50, undefined, initialEdits);
   const detail = useMessageDetail(selected);
@@ -161,8 +148,7 @@ export function MessagesView({
     ambient.set(query ? "amber" : "signal", 0.3, query ? "search" : "messages");
   }, [query, ambient]);
 
-  const searching = query.trim().length >= 2 && !semanticMode;
-  const semanticSearching = query.trim().length >= 2 && semanticMode;
+  const searching = query.trim().length >= 2;
   const list = searching ? (search.data ?? []) : (messages ?? []);
   // Discord-style order: oldest at the top, newest at the bottom. The backend
   // returns DESC (newest first); reverse so the feed reads top→bottom like DC.
@@ -283,32 +269,6 @@ export function MessagesView({
             onChange={(e) => setQuery(e.target.value)}
           />
         </div>
-        <div className="flex items-center gap-1.5 rounded-[6px] border border-hairline bg-surface-2 p-0.5">
-          <button
-            type="button"
-            onClick={() => setSemanticMode(false)}
-            className={`rounded-[4px] px-2.5 py-1 font-mono text-[10px] font-medium transition-all ${
-              !semanticMode
-                ? "bg-surface text-ink border border-hairline-focus shadow-xs"
-                : "text-ink-muted hover:text-ink"
-            }`}
-          >
-            EXACT
-          </button>
-          <button
-            type="button"
-            onClick={() => setSemanticMode(true)}
-            className={`flex items-center gap-1 rounded-[4px] px-2.5 py-1 font-mono text-[10px] font-medium transition-all ${
-              semanticMode
-                ? "bg-signal/20 text-signal border border-signal/40 shadow-xs"
-                : "text-ink-muted hover:text-ink"
-            }`}
-          >
-            <Sparkles className="size-3" />
-            SEMANTIC
-          </button>
-        </div>
-
         <div className="flex items-center gap-1.5 rounded-[6px] border border-hairline bg-surface-2 p-0.5">
           <button
             type="button"
@@ -336,76 +296,6 @@ export function MessagesView({
       </GlassPanel>
 
       <div className="grid gap-3 lg:grid-cols-5">
-        {semanticSearching && (
-          <GlassPanel className="lg:col-span-5">
-            <SectionHeader
-              eyebrow="semantic archive"
-              title={`Vector Matches for “${query}”`}
-              action={
-                <span className="mono text-xs text-[#8a8f98]">
-                  {semantic.data?.length ?? 0} matches
-                </span>
-              }
-            />
-            {semantic.isLoading ? (
-              <SkeletonRows rows={4} />
-            ) : semantic.data && semantic.data.length > 0 ? (
-              <div className="max-h-[60vh] space-y-1.5 overflow-y-auto pr-1">
-                {semantic.data.map((r, i) => (
-                  <div
-                    key={r.message_id ?? i}
-                    className="hud-card animate-stagger flex items-start gap-3 p-3"
-                    style={staggerDelay(i)}
-                  >
-                    <div className="min-w-0 flex-1">
-                      <div className="flex items-center gap-2">
-                        <span className="font-mono text-[10px] font-semibold text-signal">
-                          {(r.score * 100).toFixed(0)}% RELEVANCE
-                        </span>
-                        {r.username && (
-                          <span className="font-mono text-[10px] font-medium text-ink">
-                            {r.username}
-                          </span>
-                        )}
-                        {r.thread_name && (
-                          <span className="font-mono text-[10px] text-ink-muted">
-                            ▶ {r.thread_name}
-                          </span>
-                        )}
-                        {!r.thread_name && r.channel_name && (
-                          <span className="font-mono text-[10px] text-ink-muted">
-                            #{r.channel_name}
-                          </span>
-                        )}
-                        {!r.thread_name && !r.channel_name && r.channel_id && (
-                          <span className="font-mono text-[10px] text-ink-faint">
-                            #{r.channel_id}
-                          </span>
-                        )}
-                        <span
-                          className="ml-auto font-mono text-[10px] text-ink-muted"
-                          suppressHydrationWarning
-                        >
-                          {formatRelativeTime(r.created_at)}
-                        </span>
-                      </div>
-                      <div className="mt-1 text-xs text-ink-soft leading-relaxed">
-                        {r.content}
-                      </div>
-                    </div>
-                  </div>
-                ))}
-              </div>
-            ) : (
-              <EmptyState
-                icon={<Search className="size-7" />}
-                title="No semantic matches"
-                description="Try different phrasing — vector search inspects contextual semantics."
-              />
-            )}
-          </GlassPanel>
-        )}
-
         {/* Message Stream Deck */}
         <GlassPanel className="lg:col-span-3">
           <SectionHeader

services/frontend/src/hooks/index.ts

@@ -25,7 +25,6 @@ export {
   useRecentEdits,
   useReview,
   useReviewWsSync,
-  useSemanticSearch,
   useTextChannels,
 } from "./use-messages";
 export {

services/frontend/src/hooks/use-messages.ts

@@ -8,7 +8,6 @@ import type {
   EditHistoryRow,
   MessageActivityBucket,
   MessageRecord,
-  SemanticSearchResult,
 } from "@/lib/types";
 import type { WsHook } from "@/lib/ws-hook";
 
@@ -210,29 +209,6 @@ export function useMessageSearch(query: string, enabled: boolean) {
   );
 }
 
-// ── Semantic Search (public archive, Qdrant) ──────
-
-export function useSemanticSearch(
-  query: string,
-  enabled: boolean,
-  guildId?: string | null,
-) {
-  return useSWR<SemanticSearchResult[]>(
-    enabled && query.trim().length >= 2
-      ? ["semantic-search", query.trim(), guildId ?? ""]
-      : null,
-    async () => {
-      const res = await messagesApi.semanticSearch(
-        query.trim(),
-        10,
-        guildId ?? undefined,
-      );
-      return res.results;
-    },
-    { keepPreviousData: true },
-  );
-}
-
 // ── WS sync helpers ──────────────────────────────
 
 /**

services/frontend/src/lib/api/messages.ts

@@ -6,7 +6,6 @@ import type {
   Guild,
   MessageActivityBucket,
   MessageRecord,
-  SemanticSearchResult,
 } from "@/lib/types";
 
 export const messagesApi = {
@@ -76,17 +75,6 @@ export const messagesApi = {
       results: MessageRecord[];
     }>,
 
-  // Public semantic search over the persistent message archive (Qdrant).
-  semanticSearch: (query: string, limit?: number, guildId?: string) =>
-    orpc.messages.semanticSearch({
-      query,
-      limit,
-      ...(guildId ? { guildId } : {}),
-    }) as unknown as Promise<{
-      results: SemanticSearchResult[];
-      nextCursor: null;
-    }>,
-
   // Public, read-only activity heatmap data (per-hour volume by channel).
   getActivity: (days = 30) =>
     orpc.messages.activity({ days }) as unknown as Promise<

services/frontend/src/lib/types/message.ts

@@ -166,29 +166,9 @@ export interface AttachmentRecord {
   uploaded_at?: number | null;
 }
 
-// ── Semantic Search (read-only public archive search) ──────────
-
-export interface SemanticSearchResult {
-  message_id: string | null;
-  content: string;
-  score: number;
-  created_at: number;
-  username: string | null;
-  channel_id: string | null;
-  guild_id: string | null;
-  thread_id: string | null;
-  channel_name: string | null;
-  thread_name: string | null;
-}
-
 export interface MessageActivityBucket {
   channelId: string;
   channelName: string;
   hour: number;
   count: number;
 }
-
-export interface SemanticSearchResponse {
-  results: SemanticSearchResult[];
-  nextCursor: null;
-}