feat(archive): rich metadata in semantic search — username/channel/guild context + guild filter

Archive payload now stores username, channel_id, guild_id, thread_id and the
real message created_at (not embed time). Backend searchArray accepts an
optional guildId and applies a Qdrant payload filter so results can be scoped
to the guild being viewed. API/frontend expose the new fields and the
semantic results card shows who said it, in which channel, and when —
turning bare text blobs into contextual results. Old points fall back to
analyzed_at and omit the new fields gracefully.
This commit is contained in:
asepharyana committed 2026-09-01 18:56:14 +07:00
1 parent c704fbf7a5
commit 26a690943d
8 files changed
+84 -8

No files matched your search

@@ -18,6 +18,7 @@ export interface ArchiveMessage {
username: string;
channel_id: string;
guild_id: string;
thread_id: string | null;
created_at: number;
/** True when the message came from an age-restricted (NSFW) channel. NSFW
* content is deliberately NOT embedded into the public archive so it can't
@@ -58,6 +59,13 @@ export function archiveMessageEmbedded(message: ArchiveMessage): void {
{
text: normalized.slice(0, 4000),
flags: "",
// Rich metadata so public semantic search results can be shown in
// context (who said it, where, when) instead of a bare text blob.
username: message.username ?? "",
channel_id: message.channel_id ?? "",
guild_id: message.guild_id ?? "",
thread_id: message.thread_id ?? null,
created_at: message.created_at ?? Date.now(),
analyzed_at: Date.now(),
// 5-year persistent window (archive is NOT a TTL cache).
expires_at: Date.now() + 1000 * 60 * 60 * 24 * 365 * 5,