Fetch message history through the client REST layer

CURRENT.md

@@ -40,6 +40,13 @@ mobile draft at 390px.
   - [x] C11 preview caching: Discord-hosted image/GIF previews (WebP 320px, hash-addressed, 5 MiB
     cap, one attempt) run after every walk in both modes. Live run `...-20260916090722`:
     `previewed: 2`; thumbs served over `/api/media/<scrape>/<item>/thumb` (200).
+  - [x] C33 Puppet history pagination. The core's `message.fetch` read the client store, which
+    keeps only the newest messages plus the channel's first — every page was 10 messages and the
+    walk stopped after one page (a 10-message, 4-media scrape of an ancient group with thousands
+    of messages). It now fetches through the client's REST layer first (contiguous 100-message
+    pages, `transport: "rest"`), store as fallback. Rebuilt and restarted the managed client;
+    live re-run of channel `591000745164406784`: 21+ pages / 2,100+ messages and climbing, with
+    hundreds of images/videos being indexed.
   - [x] C32 Group-chat clarity and puppet refresh. Group DM scrapes already worked (live-verified
     against the running daemon: `zıbb & fred` and a 1:1 DM, backend `puppet`, done, previews
     cached); the Group chat tab had shown an empty list while the core was offline. It now says so

docs/core-bridge.md

@@ -68,7 +68,7 @@ Event line:
 | `dm.list` | optional `limit` | Cached DM and group DM channels with recipients, icons, and avatars. Cache-only. |
 | `message.send` | `channelId`, `content` | `{ channelId, messageId, sent: true }` |
 | `message.read` | `channelId`, `limit` | `{ cached, messages }`. Cache-only. Messages carry the author avatar, attachment metadata (id, filename, URL, content type, size, dimensions, duration), embeds (video, image, and thumbnail URLs plus their dimensions), sticker items, forwarded snapshots, a voice-message flag, `editedAt`, and `replyTo`. |
-| `message.fetch` | `channelId`, `limit`, optional `before` or `after` | `{ fetched, messages }`. Fetches from Discord. Same message shape. |
+| `message.fetch` | `channelId`, `limit`, optional `before` or `after` | `{ fetched, transport, messages }`. Fetches from Discord through the client's REST layer, so pages are full and contiguous; the client store (which keeps only the newest messages plus the channel's first) is the fallback. `transport` is `rest` or `store`. Same message shape. |
 | `reaction.add` | `channelId`, `messageId`, `emoji` | `{ added: true }` |
 | `dm.send` | `userId`, `content` | `{ userId, channelId, sent: true }` |
 

docs/media-scraper.md

@@ -18,7 +18,7 @@ interface ScrapeBackend {
 | Backend | Implementation |
 | --- | --- |
 | `slopbot` | discord.js REST (`channels.messages.fetch`), bucket-aware client |
-| `puppet` | The core action `message.fetch` over the bridge. Expired URLs refresh with `before: id+1`, which returns exactly the target message |
+| `puppet` | The core action `message.fetch` over the bridge: the client's REST layer paginates (100/page, contiguous), with the client store as fallback. Expired URLs refresh with `before: id+1`, which returns exactly the target message |
 
 The daemon picks `slopbot` when the channel is a guild channel and slopbot is configured and in the
 guild; otherwise `puppet`. A scrape records the backend it used. The walk can resume on the same

src/renderer/puppet/actions.ts

@@ -23,6 +23,7 @@ import {
     MessageStore,
     PresenceStore,
     ReadStateStore,
+    RestAPI,
     UserStore,
     VoiceStateStore
 } from "@vencord/types/webpack/common";
@@ -961,6 +962,47 @@ export async function handlePuppetAction(action: string, params: Record<string,
                 const before = params.before as string | undefined;
                 const after = params.after as string | undefined;
 
+                const inRange = (message: any) => {
+                    if (after && compareIds(message.id, after) <= 0) return false;
+                    if (before && compareIds(message.id, before) >= 0) return false;
+                    return true;
+                };
+
+                const normalize = (list: any[]) => {
+                    const seen = new Set<string>();
+                    return list
+                        .filter(message => inRange(message) && !seen.has(message.id) && seen.add(message.id))
+                        .sort((a, b) => compareIds(a.id, b.id))
+                        .slice(-Number(limit));
+                };
+
+                // Preferred: the client's REST layer paginates properly. The store keeps only
+                // the newest messages plus the channel's first, so reading it back yields tiny,
+                // non-contiguous pages.
+                let messages: any[] | null = null;
+                let transport = "rest";
+
+                try {
+                    const query: Record<string, string | number> = { limit: Number(limit) };
+                    if (before) query.before = before;
+                    if (after) query.after = after;
+
+                    const response = await RestAPI.get({
+                        url: `/channels/${channelId}/messages`,
+                        query: { ...query }
+                    });
+                    const list = Array.isArray(response)
+                        ? response
+                        : Array.isArray((response as any)?.body)
+                          ? (response as any).body
+                          : null;
+                    if (list) messages = normalize(list);
+                } catch {
+                    messages = null;
+                }
+
+                if (messages === null) {
+                    transport = "store";
                     await fetchMessages({
                         channelId,
                         limit: Number(limit),
@@ -969,16 +1011,14 @@ export async function handlePuppetAction(action: string, params: Record<string,
                     });
 
                     const all = MessageStore.getMessages(channelId)?._array ?? [];
-                const filtered = all.filter(message => {
-                    if (after && compareIds(message.id, after) <= 0) return false;
-                    if (before && compareIds(message.id, before) >= 0) return false;
-                    return true;
-                });
+                    messages = normalize(all);
+                }
 
                 return ok({
                     channelId,
-                    fetched: filtered.length,
-                    messages: filtered.slice(-Number(limit)).map(mapMessage)
+                    fetched: messages.length,
+                    transport,
+                    messages: messages.map(mapMessage)
                 });
             }