recovery-worker used a hardcoded STUCK_PROCESSING_AGE_MS=300_000 while
messagesCleanup's default and AI_ANALYSIS_PROCESSING_TIMEOUT_MS are both
120s. Rows stuck between 2 and 5 minutes were never reverted by the
recovery worker — they looked permanently stuck (and accumulated under
load) even though the batch budget had long passed. Now derives the
threshold from config so the two knobs can never drift again.
Piscina worker threads outlive process.exit() and linger as orphaned
processes holding DB connections/locks after a deploy restart. Two live
gateways then fight over the same messages table rows (one claims
processing, the other reverts), which left messages stuck in
ai_status='processing' forever.
Destroy both worker pools before closing the DB, with a 5s fallback so
a hung vision job cannot block shutdown indefinitely.
getPendingMessagesByConversation() flips every fetched pending row to
'processing', then pickBatchWithinBudget() may stop early on the token
budget. The tail rows that did NOT make the batch were never un-claimed,
so they stayed 'processing' forever — the recovery worker reverted them
(120s) only for the next wave to re-claim them, an infinite loop of
stuck messages that never get analyzed (saw 22 rows, some recycled for
40+ minutes).
- add computeBudgetOverflowMessages() pure helper (batchBudget.ts)
- batchScheduler un-claims overflow rows back to 'pending' before
dispatching the trimmed batch
- recovery-worker now reverts stuck processing unconditionally (the old
'conversationProcessing.size > 0' guard skipped the revert when the
in-memory lock map was empty, e.g. fresh boot — exactly when stranded
rows from a previous process need rescuing)
- 3 regression tests for the overflow helper
GMW moves off omniroute (100.121.180.82:20128) and off the direct NVIDIA
vision endpoint onto 9router, which runs on the same host as both services
(127.0.0.1:4014) — loopback avoids the TLS/proxy hop and localhost calls
bypass 9router's remote-key guard.
- gateway + backend: AI_LLM_BASE_URL default -> http://127.0.0.1:4014/v1
- drop stale 'omniroute' router references from comments/docs now that the
active router is 9router (llmClient, llmCaller, ARCHITECTURE, AGENTS)
Verified against 9router before wiring: model 'text' -> gemini-3.5-flash-lite
(SSE, as the pipeline expects), 'multimodal' -> nemotron-3-nano-omni answers
image input, and gemini/gemini-embedding-001 returns 3072 dims — matching the
existing Qdrant collections (no reindex needed). The GMW key is already
registered in 9router's apiKeys table.
typecheck + lint + tests green (gateway 138, backend 37 excluding e2e).
The metrics server binds METRICS_PORT (code default 9090; this host runs it
on 4018 — 4016 is occupied by another process). Docs said 4016 in three
places, which is not what the service does.
README.md was the extraction-era document (referenced winston, mock-crc.ts,
llmModerationClient.ts, indonesianTextNormalizer.ts — all long gone) and
duplicated ARCHITECTURE.md. Rewritten as a short run-the-service guide;
layout/design lives only in ARCHITECTURE.md.
MODULE_STRUCTURE.md deleted: it was a stale duplicate of ARCHITECTURE.md,
referenced by nothing but itself.
ARCHITECTURE.md updated to the post-refactor reality: app/ lifecycle split
(bootstrap/lifecycle/process-guards/metrics-collector), ai-moderation
recovery-worker + cache-prune, per-module index.ts facades, one-way
dependency rule, corrected init/shutdown/observability sections.
app/:
- bootstrap.ts 277 -> 145 lines: config guard, DB connect, client debug
logging and startup order are now named steps with a comment header
- lifecycle.ts (new): everything wired on the Discord 'ready' hook, in
explicit order (inject broadcaster -> register listeners -> start workers)
- process-guards.ts (new): SIGINT/SIGTERM/uncaughtException/unhandledRejection
in ONE place, using isTransientStreamError() instead of two duplicated
inline code lists
- metrics-collector.ts (new): AI pipeline Prometheus gauges
modules/:
- ai-moderation/index.ts + message-capture/index.ts (new): public facades so
app/ never reaches into internal files
- aiAnalyzer.ts 317 -> 146 lines: pure entry API; skip-verdict recording
extracted into recordSkip()
- recovery-worker.ts (new): stranded-message recovery + stale lane/CB pruning
- cache-prune.ts (new): 6h expired-verdict sweep, throttled + resettable
- drop 3 dead re-exports (pickBatchWithinBudget/onCircuitBreakerAlert/
getConversationKey) whose consumers import the origin files directly
No behavior change. typecheck + lint + 138 tests green; nix build OK.
- delete src/shared/index.ts fat barrel; point 10 importers at the exact
module they use (redis-channels, moderation-types, utils/pagination)
- message-capture/types.ts re-exports from shared/moderation-types directly
- shared/errors: add errorMessage() + isTransientStreamError() helpers,
replacing the repeated err-message and transient-code checks
- drop unused imports flagged by biome
No behavior change. typecheck + lint + 138 tests green.
Image messages previously blocked the whole analysis pipeline:
- conversationProcessing was a single lock per conversation; processBatch
awaited BOTH text and media worker jobs before releasing it, so a fast
text verdict sat unused until the slow vision/media batch finished
- one global LLM semaphore (AI_LLM_MAX_CONCURRENT) was shared by text and
media, so a vision backlog could starve text inference
- recovery worker gated on conversationProcessing.size
Now the queue is split into independent text/media lanes:
- conversationProcessing maps key -> Partial<Record<lane, startedAt>>;
each lane holds its own lock and frees it the moment ITS worker job
resolves (ownership-guarded clear prevents stale timers clearing newer
slots)
- two LLM semaphores: AI_LLM_MAX_CONCURRENT (text, default 8) and
AI_LLM_MEDIA_MAX_CONCURRENT (media, default 4) via
withLlmConcurrency(fn, { lane })
- batchScheduler schedules per conversation+lane (timer keys
'<key>::<lane>'); splitMessagesByLane/laneOfMessage moved to pure
analysisLanes.ts (unit-testable without Piscina)
- ai-analysis-worker batch jobs carry a lane field; per-lane active
request gauges (active_text_requests / active_media_requests)
- added tests/analysisLaneLock.test.ts (7 tests: independent lane locks,
preserving other-lane lock, clear-all, ownership guard, lane split)
Docs: ARCHITECTURE.md + AGENTS.md concurrency model updated.
typecheck/lint/test(138)/build all green.
Jev (oc/jev-1.13-free via 9router /v1/systemone) added as primary text
analyzer was underperforming. Delete the whole feature:
- jevAnalyzer.ts + its unit & live-smoke tests
- Jev-first branch in textBatchProcessor, restore pure callModerationLLM path
- AI_LLM_JEV_* config vars (zod) and .env.example entries
- @typesafe-ai/sdk dependency (+ lockfile)
Behavior: text moderation is LLM-only again, exactly as before the
Jev feature; AGENTS.md invariant 'LLM is the only judge' holds.
Show relative time, always-on username with channel context, and
collapse top emojis to two; add VIEW ALL toggle to reveal all 20
fetched reactions instead of the top 4.
* feat(gateway): Jev (System One) as primary text moderator with LLM fallback
Add TypeSafe Jev via @typesafe-ai/sdk v0.6.0 as the PRIMARY analyzer for
text-only moderation sub-batches; the existing LLM stays as the fallback
for anything Jev cannot decide confidently (per-message gate rejection or
API failure) and for media batches.
- jevAnalyzer.ts: TypeSafeClient wrapper, declarative state builder
(System One models MUST get factual state, not chat-XML — chat framing
made Jev confidently wrong on clean messages at 0.98 confidence),
message-id-keyed question builder (5 typed questions per message),
cross-consistency acceptance gate (noul↔status↔severity↔action↔category),
answer→AnalysisResult mapper, fail-open outcome.
- textBatchProcessor.ts: Jev-first per sub-batch, rejected ids + API
failures fall back to callModerationLLM; no cross-batch pollution.
- config: AI_LLM_JEV_ENABLED/API_KEY/BASE_URL/MODEL/TIMEOUT_MS/MIN_CONFIDENCE.
- .env.example: documented all 6 Jev env vars.
- tests: unit (question/state builders, gate, mapper) + live smoke
(gated behind AI_LLM_JEV_SMOKE=1) verified 4/4 accepted vs real 9router.
* fix: auto-fix code quality [skip ci]
Add tests/llmE2e.test.ts — 7 end-to-end tests driving the REAL
moderation prompt pipeline (buildSystemPrompt → XML payload → llmChat
→ parseModerationResponse) against a live model via omniroute.
Covers: clean technical content (no false positives), harassment
(flagged), username-only offenses including 'Pecinta Pria' +
sexual/provocative usernames + SARA-in-username (always warn/low,
NEVER delete — the nickname-reset path), and spam bursts.
Gated behind AI_LLM_BASE_URL + AI_LLM_API_KEY: CI (no creds) skips
the file → 216 unit tests stay green, zero LLM cost. Run locally via
pnpm test:e2e:live (scripts/run-llm-e2e.sh injects creds from bws).
Verified: 223/223 tests pass with live LLM, stability across 4 runs,
typecheck + biome clean. docs: TESTING.md. ignore .hermes/ plans.
- rules.ts: explicit rule that sexual/provocative username terms
(Pecinta Pria, Cinta, pacar, janda, bokep, hot, seks, nude, telanjang)
MUST be flagged as offensive_username with low severity
- output.ts: output schema case for sexual/provocative username
violations, always status: warn, severity: low, never flagged/delete
No hardcoded keyword lists — fix is at the LLM prompt level only.
- new tinyFishSearch module: GET api.search.tinyfish.ai with X-API-Key,
maps top-3 to SearchResult shape, never throws (all failure modes -> [])
- wikipediaSearch: on wiki miss, one tinyfish attempt; hits cached 6h
under the same key so fallback latency is paid once
- termGlossary: on summary miss, top tinyfish hit becomes the definition
(persisted permanently like wiki defs); miss keeps 1h sentinel
- config: TINYFISH_API_KEY (empty = fallback disabled), ENABLED,
BASE_URL, TIMEOUT_MS, LOCATION, LANGUAGE knobs
- tests: 6 coverage for disabled/mapping/non-OK/network/bad-json
API key NOT committed — set TINYFISH_API_KEY in BWS gmw secrets.
Verified: typecheck + lint clean, 216/216 tests pass, live probe
'gubernur jawa barat' returned 3 mapped results
- attachmentUploader: downloadDiscordAttachment retries CDN timeouts
via retryWithBackoff (ATTACHMENT_RETRY_ATTEMPTS, 1s-8s backoff);
AbortError normalized so 403/404 refresh path never fires on timeouts
- attachmentUploader: uploadAttachmentToTele uses ATTACHMENT_RETRY_ATTEMPTS
instead of retries:0 (Tele 5xx under load was failing outright)
- wikipediaClient: wikipediaSummary retries once on abort/timeout
(100% of prod summary errors were aborts); termGlossary drops its
redundant second call (was up to 4 reqs/term under miss+retry)
Verified: typecheck + lint clean, 210/210 tests pass
- Added eager selfbot voice connection establishment to ensure video capture works seamlessly during voice channel joins.
- Introduced a manual video watch command for selfbots to allow operators to initiate screen recording of other members' streams.
- Enhanced video recording functionality to split recordings into segments based on user activity, similar to voice recordings.
- Fixed mobile navigation issues by extending the navbar to include all items and ensuring responsive form controls.
- Standardized error handling across frontend components to improve user experience during failures.
Jockie Music (user 411916947773587456) posts now-playing embeds/spotify links
~1347 captured messages — every one consumed a moderation LLM call for zero
signal and contributed to batch timeouts. Config AI_SKIP_ANALYSIS_USER_IDS
(default=Jockie) skips them at ALL three analysis paths:
- queueMessageAnalysis entry (direct skip-result like age-restricted)
- batchScheduler processing (pre-batch filter)
- individual recovery path (no fallback spam for already-skipped authors)
Skip-result mirrors age_restricted: status=clean, flags=[skip_analysis_user],
action=none — stays visible in the dashboard, never analyzed.
The text model behind omniroute/9router consistently takes >45s on long-context
batches. At 45s every such batch fell through to the individual-fallback
queue which re-runs with its own timeout, then exhausted to ai_status=error.
75s keeps the bounded budget while letting the first-pass batch succeed.
- AI_LLM_MEDIA_ANALYSIS_TIMEOUT_MS 60s→120s + vision 60s→120s: vision model
via router regularly exceeded 60s, dropping media batches into the
individual-fallback chain then exhausting into ai_status=error.
- Qdrant upserts: retryWithRetry() wraps PUT /points with exponential
backoff (3 attempts, jitter) for transient 408/abort/ECONNRESET — the
41 six-hour 'Qdrant upsert failed — semantic entry skipped' warnings were
single-hop timeouts on a healthy-but-loaded Qdrant.
- LLM caller: on parse failure, attempt extractJson() structural repair of
the raw content (models with thinking disabled sometimes emit JSON as
plain text) before giving up and re-requesting.
C2 full scope persistence:
- getRecentConversationContext now returns per-turn guildId/channelId
- processMessage merges historical scope when current request is unscoped
- chatbot remembers server context across all 8 history exchanges (not just 3)
A4+Metrics:
- Add incrementCounterBy(name, delta, labels) to gateway-metrics
- Token-usage counters: llm_tokens_total{model, type, label} per batch
- Cache hit counters: moderation_cache_hits{type} exact/semantic-qdrant/semantic-pg
- Cache miss counters: moderation_cache_misses per batch
- Prometheus /metrics now exposes cost + cache hit-rate for dashboards
Performance:
- Parallel tool execution within each chatbot round (Promise.all)
- All tool results collected before sending to model (ordering preserved)
- Tool failure now logged with structured warning (chatbot.tools.ts)
Verified: backend tsc+biome 37/37, discord-gateway tsc+biome 210/210
- Send system prompt as role:system instead of merging into the user message
(models treat system messages as authoritative -> better tool selection)
- Add full 14-tool catalog + multi-hop instruction to the system prompt
- Raise MAX_TOOL_ROUNDS 4->6 so multi-hop queries complete
- Raise conversation history 3->8 exchanges (less repeat questioning)
- Append guild/channel scope hint to each user turn
- Add saturating toolErrors counter + structured warn log per tool failure
so chatbot tool failures are observable instead of silently absorbed
Even with the prompt firewall (username vs content), the LLM can still
occasionally mis-apply a content-level zero-tolerance flag (sara /
conflict_instigation) to a message whose ONLY violation is the username
(e.g. 'matikanetanyahu'). The auto-delete eligibility check only recognized
exact offensive_username flags, so such false positives still deleted the
message.
Add a belt-and-suspenders guard in isNicknameOnlyViolation: if the flag set
is entirely username-attributable (offensive_username/sara/conflict_instigation)
AND the analysis text corroborates that the violation is username-only with
clean message content, route to nickname-reset instead of message deletion.
Adds 6 test cases covering the real matikanetanyahu scenario and the
false-positive/negative boundaries.
Add explicit FIREWALL PENILAIAN rule separating username assessment from
message content assessment. Username containing political/religious terms
(matikanetanyahu etc) is assessed as offensive_username with severity low
— never triggers sara/conflict_instigation zero tolerance for content.
Changes:
- rules.ts: Add FIREWALL section before LARANGAN BERAT; clarify each
zero-tolerance rule applies to ISI PESAN only; update hierarchy #9
- examples.ts: Fix example #11 status flagged→warn for clean username;
add example #11b with matikanetanyahu case
- output.ts: Explicit status/action mapping for username-only violations
result.score was required by zod; the LLM (gemini-3.5-flash-lite via 9router)
occasionally omits it for media batches, hard-failing the whole batch parse
('Zod validation failed: expected number, received undefined' at
results[0].score). Callers already null-coalesce (result.score ?? 0) and the
parser clampScore()s it, so requiring it only caused parse failures.
Adds regression tests: media-batch without score parses (score->0), and
score-present responses still parse with the value.
A selfbot (user token) cannot auto-detect other members' camera/share
(no VOICE_STATE_UPDATE for others, 403 on member fetch). The only
selfbot-viable path to capture another member's SCREEN SHARE is an
operator-initiated STREAM_WATCH (gateway op 20, not gated on bot-vs-user).
Add video:watch / video:unwatch Redis commands routed via the existing
command handler to startStreamWatch/stopStreamWatch, which then does the
DAVE handshake + per-burst MP4 segmentation + DB insert + Tele upload
(already implemented in streamWatchReceiver).
- new VideoHandler (command-handler/video.handler.ts)
- register video:watch / video:unwatch in handler-registry + CommandHandler
- command constants COMMAND_VIDEO_WATCH / COMMAND_VIDEO_UNWATCH
- resolve active voice channel from voice controller + client cache
- 8 unit tests (videoHandler.test.ts)
- biome fixes for pre-existing test import ordering
All green: typecheck, build, lint (174 files), 200 tests.
Dump all WS raw event types after 10s (see if GUILD_CREATE exists at all),
and log broadcaster_user_ids + sessions.voice from READY payload.
Selfbot may RESUME (skip GUILD_CREATE) — voice data may only be in READY.
Temporary diagnostic to see whether GUILD_CREATE.voice_states actually
reaches the selfbot raw listener (selfbot-v13 emits everything via
WebSocketShard Events.RAW). Will remove once root cause confirmed.
Real root cause of 'tidak mendeteksi user yg sudah ada di voice':
The selfbot's discord.js-selfbot-v13 GUILD_CREATE handler only sends
GUILD_SUBSCRIPTIONS_BULK — it DROPS d.voice_states from the payload.
So when the gateway (re)starts, every user who was ALREADY in the voice
channel (with camera or screen-share on) is invisible: channel.members is
empty, and REST fallbacks don't work for user tokens (verified live:
GET /channels/{id}/voice-states -> 404, GET /guilds/{id}/members -> 403).
Fix: register our own raw listener for GUILD_CREATE, capture
d.voice_states per guild (buffer it), and consume it in
scanExistingStreamers when the channel gets tracked (voice join happens
after GUILD_CREATE). This is the ONLY user-token-compatible source of
'who is in voice with video right now'.
- videoRecorder: +pendingGuildCreateVoiceStates buffer, +raw GUILD_CREATE
listener, Path C consumes buffered states (self_video/self_stream,
matching channel, skip self) before the REST fallback
- scanExistingStreamers: Path A cache -> Path C GUILD_CREATE -> Path D
REST members.fetch() best-effort
- +1 test: GUILD_CREATE voice_states buffer -> watch camera+screen-share,
skip self/no-video/other-channel (192/192 pass)
- typecheck + build + biome clean (1 pre-existing warning)
Root cause: scanExistingStreamers only read channel.members, which is
EMPTY after a gateway restart because the selfbot's guild member cache
hasn't been populated yet. So any user who was ALREADY on camera /
screen-sharing when the bot (re)joined was never detected → no video.
Fix: when the member cache is empty, fall back to guild.members.fetch()
(REST GET /guilds/{id}/members — user-token compatible; the bot-only
GET /channels/{id}/voice-states returns 404 for selfbots, verified live)
which populates member.voice states, then re-scan channel.members for
streaming/selfVideo.
- scanExistingStreamers is now async; trackChannel fire-and-forgets it
- logs source=cache vs source=rest-members for observability
- +1 test: cold-start channel.members empty → REST fetch → watch camera user
- 191/191 tests pass, typecheck + build + biome clean
Previously forceSelfServerUnmuteUndeafen only ran on video-watch attempts and
after voice reconnects — so when an admin server-muted or server-deafened the
bot, it stayed muted/deafened for minutes (or forever if no streamer came on).
Add registerSelfVoiceStateGuard: a voiceStateUpdate listener that detects the
bot's own serverMute/serverDeaf transition to true and immediately re-issues
mute:false,deaf:false. Wired at client-ready in bootstrap.ts. Best-effort,
idempotent, never blocks the gateway.
Closes the three highest-value gaps from the FE data-flow audit:
- voice/view.tsx: render VoiceStatus.connections (multi-guild bridge roster)
which was fetched but never shown — channel name, guild/channel id,
connected time, ACTIVE marker for the primary bridge (falls back to the
active channel when connections is empty).
- recordings/view.tsx: show SpeakerSummary.last_at ('active <relative>') on
the speaker leaderboard row (data was fetched but hidden).
- dashboard/view.tsx: render activity.hourly as an 'Hourly Flow · Last 24h'
24-bar volume strip with per-hour tooltip and total msgs/flagged header
(hourly distribution was fetched but only daily was charted).
Verified: tsc clean, biome clean (127 files), pnpm build green.
Follow-up to d0453881: the pgUserReputationsTable definition and the
UserReputation/UserReputationInsert type aliases were the only remaining
references to the removed user reputation feature. Nothing imports them
(verified via grep across src/ + tests/), so they're pure dead code that
could mislead future devs into re-joining a dropped table. Railed out to
close the 42P01 failure class completely.
Root cause: gateway migration 0016 removed the per-user reputation feature
(DROP TABLE user_reputations), but the backend still LEFT JOINed
pgUserReputationsTable in dashboard listUsers/getUserDetail and the chatbot
get_user_reputation tool. Against the live DB these queries raised 42P01
(undefined_table) -> the new /users page could never load (oRPC WS error).
Fix:
- dashboard.repository.ts: remove reputation columns/join from listUsers and
getUserDetail (data sourced only from messages + user_profiles).
- chatbot.tools.ts: get_user_reputation now returns honest 'unavailable'
(feature removed) instead of querying the dropped table.
- chatbot.toolDefs.ts: update tool description so the LLM doesn't advertise
a dead feature.
- frontend types/users view: drop trust_score/clean_message_streak/
total_infractions/last_infraction_at; derive risk label from real flagged%
and add warn_count to the inspector (real available data).
Verified: backend+frontend tsc clean, biome clean, 37 unit tests pass,
query tested against live DB (returns real members), build green.
videoReceiver.getSsrcInternalMap read asAny._map, but @discordjs/voice
0.19.x exposes the SSRC map as the public field 'map'. So inferVideoOwner
(e.g. matching a video RTP SSRC to a user via audio-SSRC proximity) always
returned null -> every video packet was silently dropped after decryption.
Accept both 'map' and '_map'. Unblocks camera/screen-share attribute when
op12 does not carry a videoSSRC stream entry.
- New /users route: member roster (trust score, clean/flagged counts, message volume)
+ inspector detail (trust breakdown, clean streak, infractions, AI profile,
recent messages) with live search
- Backend dashboard.listUsers/getUserDetail now JOIN user_reputations to expose
trust_score, clean_message_streak, total_infractions, last_infraction_at +
warn_count/clean_count breakdowns
- Frontend types updated to match; useUsers refactored to PaginatedUsers shape,
added useUserSearch for the old search behavior
- Nav rail + mobile nav: add Users entry
streamWatchReceiver: DAVE decrypt for screen-share video packets was
returning null every time (VIDEO-PKT diag showed packets arriving but
no 'Video segment opened'). Three possible causes:
1. No VIDEO decryptor in MLS group (audio-only handshake)
2. GoLive stream tags video packets as AUDIO
3. Screen-share payloads are unencrypted above the AES layer
Fix: try MediaType.VIDEO first, then MediaType.AUDIO, then passthrough
(legacy-decrypted payload as-is). This covers all three modes without
breaking audio recording.
Fork @discordjs/voice 0.19.2 into vendor/discord-voice-fork and patch the
voice gateway handshake so Discord sends camera/screen-share RTP video:
- Identify payload now declares video:true + streams:[] (derived from
Discord-RE/Discord-video-stream) - this is what makes Discord deliver
H264 (PT 96-127) to the voice socket. Previously the client never
declared video capability, so Discord omitted all video SSRCs/packets.
- op-12 Speaking handler maps streams[].ssrc -> videoSSRC alongside the
audio SSRC, so ssrcMap carries camera/screen-share stream IDs.
- SSRCMap.get() now also resolves video SSRCs (video RTP arrives on a
different SSRC than audio), enabling attribution for videoReceiver.ts.
- Both dist/index.js (CJS) and dist/index.mjs (ESM) patched; 3 isolated
hunks vs upstream, verified by diff.
- package.json points @discordjs/voice -> file:vendor/discord-voice-fork
(pnpm lockfile updated, CI --frozen-lockfile compatible).
- .gitignore: replace global dist/ with per-service explicit patterns so
the vendored fork's dist/ is committed while build outputs stay ignored.
- tests: +3 SSRCMap videoSSRC cases (190 total, all pass).
The receive-side pipeline (H264Depacketizer -> .h264 -> muxToMp4 -> MP4)
already exists in videoReceiver.ts; this unblocks it by making Discord
actually deliver video packets.
Previously only member.voice.streaming (screen share, self_stream) triggered
video capture. Discord reports camera via self_video:true, so camera-only
users were never watched. Now scanExistingStreamers and handleVoiceStateUpdate
start a watch when streaming OR selfVideo is set, and stop it when both clear.
Video (camera + screen share) DAVE stream-watch now produces per-burst
MP4 segments instead of one long .h264 per watch:
- Detects VIDEO_SILENCE_MS (4000ms) of no H264 packets → closes the
current segment, muxes to MP4, registers in voice_recordings + uploads
to TeleUploader, then reopens for the next burst (mirrors voice AfterSilence).
- Per-watch segment counter + per-segment depacketizer reset + closing
guard + write-error swallow so races (silence close vs in-flight UDP
packet) never corrupt files or crash the gateway.
- Frontend: recordings deck renders a native <video> player for MP4 rows
(detected by filename), keeps single-playback registry across audio+video.
Gateway archive embedder now parses metadata.channel.{channelName,threadName}
from each message and stores channel_name/thread_name in the Qdrant payload.
Backend exposes them; the semantic results card renders the thread name (or
channel name) instead of a raw #snowflake, with the ID as a last-resort
fallback for legacy points. Matches the message feed's channel-label logic
(getMessageChannelLabel).
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.
- AI_VOICE_TRANSCRIPTION_MODEL config (default whisper-1) so the model can be a provider-qualified id (openrouter/openai/whisper-1) that actually has credentials through 9router/omniroute — bare whisper-1 maps to the openai provider which has none
- response_format json (not text): 9router proxies only json/verbose_json transcription responses; text returns 400
- parse text from the json response object
- prod env updated: model=openrouter/openai/whisper-1 (still needs OpenRouter STT balance — 402 until funded)
- Normalize text before embedding (strip mentions/URLs/emoji/markdown/control chars, lowercase, truncate) on both write and query sides so vectors aren't diluted and tokens aren't wasted
- embeddingClient: retry embeddings (maxRetries 2), validate batch dimension consistency, preserve index alignment for empty-normalized texts
- archiveEmbedder: store normalized text in archive payload, skip empty-normalized content
- backend: normalize search queries, make archive search similarity threshold configurable (AI_LLM_EMBEDDING_ARCHIVE_MIN_SIMILARITY, default 0.6)
The bot's own VOICE_STATE_UPDATE showed server-level deaf:true — a
server-deafened member is NOT sent the streamer's audiovisual RTP by Discord,
which is the likely reason no H264 arrives despite the DAVE watch reaching
Ready. The previous fire-and-forget forceSelfServerUnmuteUndeafen() ran once
after the first Ready join and silently reverted on reconnect/restart.
- Export forceSelfServerUnmuteUndeafen from recorder.ts; re-assert it (with
read-back verification logging stillDeaf) at the START of every
startStreamWatch() before STREAM_WATCH is sent (dynamic import avoids the
recorder <-> videoRecorder <-> streamWatchReceiver module cycle).
- Re-assert it again after a successful voice reconnect.
- startStreamWatch() is now async; callers use void.
maxLen stayed 72 across 243 packets (no real H264, which is hundreds+ bytes) —
only 44-72-byte RTP packets on PT 76/72/73 arrive. Add ssrc + first-32-bytes
hex so we can identify exactly what Discord sends to the watch socket (control
packets vs stale video), which determines whether the gap is upstream routing
or whether large H264 packets are missing entirely.
Enhance watch-socket diagnostic to report distinct RTP payload types seen and
the max packet length, so we can distinguish 'only small control packets arrive
(no real H264)' from 'H264 arrives but decrypt fails'. Live already confirmed
dave=true ready=true with packets flowing but no burst — need to know if they're
tiny 52-byte control packets (PT 73) or large H264.
Instrument handleUdpMessage to log (rate-limited, first 3 then /20s) whether
video RTP packets actually reach the watch socket, and whether the DAVE session
is attached+ready and encryption key material present. Needed to diagnose why
no .h264 is written despite DAVE Ready + MLS: is the packet not arriving, or is
decrypt returning null?
CRITICAL: djs/voice stores the DAVESession wrapper at net.state.dave
(createDaveSession assigns to state.dave on op4 SessionDescription), NOT
inside connectionData. decryptVideoPacket looked up connectionData.dave which
is ALWAYS undefined -> every video packet hit '!dave?.session' guard and was
silently dropped, so no .h264/.mp4 ever got written despite the handshake
reaching Ready.
Fix: pass net.state.dave as a separate arg (the wrapper has .session ->
Davey.DAVESession) so the DAVE-layer decrypt (MediaType.VIDEO) actually runs.
Typecheck + build pass, lint clean (src/), 179/179 tests.
Live deploy 15:51 reached DAVE watch READY + completed MLS handshake on the
stream RTC (0->1->2->3->4, MLS commit processed, heartbeats alive). scan-on-
join also confirmed: 'Scanned pre-existing streamers on join watched=1'.
P4 remaining = capture actual video RTP (burst->mp4) while a stream is live.
If someone is ALREADY sharing screen / camera on when the bot joins the
channel, no voiceStateUpdate with streaming:true fires for them, so the
bot never sent STREAM_WATCH and missed their video entirely. trackChannel
now scans channel.members and starts a watch for anyone already streaming
(ignoring the bot itself and non-streamers). Idempotent: startStreamWatch
no-ops if a watch already exists. +2 tests (9/9 in videoRecorder).
djs/voice Networking emits stateChange(oldState, newState), but the watch
handler declared (newState, oldState) -- reversed. So the code-4 Ready
branch (which attaches udp.on('message') + logs 'DAVE watch READY') never
fired when entering Ready; it only fired spuriously when LEAVING Ready.
Result: full DAVE handshake completed on the watch RTC (Ready + DAVE MLA +
video stream 21029 active 1920x1080@60) but no UDP listener -> no video
captured. Swap to (oldState, newState) so enter-Ready wires the socket.
Add watch-state N->M log on every djs/voice Networking stateChange (with
hasUdp flag) so the stream-watch connection's exact progression is visible:
OpeningWs(0)->Identifying(1)->UdpHandshaking(2)->SelectingProtocol(3)->
Ready(4). Pins down where the DAVE flow stalls instead of guessing from the
absence of logs. Pairs with debug:true + watch-djs-debug.
Add debug:true to watch Networking options and wire net.on('debug') to
logger.info so djs/voice internal WS/DAVE state transitions appear in
journal. Without this, the stream-watch connection went silent after
'Streaming DAVE Networking' — no ready/error/close visible. Needed to
diagnose why the WS to stream endpoint 'c-sin14-xxx:2083' produced no
events.
Live log (14:52) showed the stream-watch flow reaching STREAM_CREATE +
STREAM_SERVER_UPDATE but then DAVE processProposals threw
'ValidationError(WrongGroupId)' -- the Davey MLS session derived the wrong
group because connectionOptions.channelId was the guild voice channel id.
Per Discord-RE StreamConnection.daveChannelId = BigInt(serverId)-1n, the
stream-watch DAVE MLS group is keyed to rtc_server_id-1, not the vc channel.
Fix: pass BigInt(serverId)-1n as channelId to the watch Networking.
This error also surface as an uncaughtException that crashed the gateway
(systemd restarted it). Correct channel id prevents it at the root.
The previous code read the sessionId from the selfbot client's voice manager
(client.voice.connection), which is no longer established since ensureSelfbotVoice
was removed — it would have sent sessionId:'none' in the watch Networking
identify and been rejected. Read the active session from the guild
@discordjs/voice connection (getVoiceConnection(guildId).state.networking...
connectionOptions.sessionId) instead.
The first streamWatchReceiver only did dave.session.decrypt(msg.subarray(12))
which skipped the outer legacy-AES layer Discord wraps around the DAVE payload
on every RTC packet (encrypt = dave.encrypt then aead_aes256_gcm with RTP
header as AAD). Port @discordjs/voice VoiceReceiver.decrypt/parsePacket
faithfully (header strip incl CSRC+extension+padding, AES-GCM auth tag, then
DAVE MediaType.VIDEO). Without this the .h264 would be garbage.
Replace the dead selfbot-v13 video path (WS 4017 DAVE). videoRecorder
now delegates to a new streamWatchReceiver that:
- sends STREAM_WATCH (op 20) on voiceState.streaming
- opens a @discordjs/voice Networking to the watch RTC (STREAM_CREATE +
STREAM_SERVER_UPDATE) with DAVE enabled
- decrypts H264 via Davey MediaType.VIDEO, depacketizes + muxes to mp4
- tears down on streaming-stop / leave / untrack
Remove ensureSelfbotVoice/createVideoStream/joinStreamConnection (dead).
recorder.ts no longer fires the futile eager selfbot join.
Supersedes the eager-selfbot connection plan: Discord now REQUIRES DAVE (E2EE,
WS 4017) on all voice RTC, and discord.js-selfbot-v13's voice stack predates
DAVE, so its video-receive path (joinChannel + joinStreamConnection +
receiver.createVideoStream) cannot authenticate. @discordjs/voice 0.19.2 exports
VoiceWebSocket/VoiceUDPSocket/DAVESession/Networking + @snazzah/davey supports
MediaType.VIDEO/Codec.H264 decrypt, so we can build a DAVE-capable stream-watch
connection. Phased plan: prototype (P2), gateway integration (P3), live verify (P4).
Video capture (camera + screen share) recorded ZERO frames because the selfbot
ClientVoiceManager.connection was created LAZILY — only when a user started
streaming. At that point the bot is already connected via @discordjs/voice, so
the selfbot re-join never gets a fresh VOICE_SERVER_UPDATE and times out with
VOICE_CONNECTION_TIMEOUT after 15s. joinStreamConnection (STREAM_WATCH) +
receiver.createVideoStream both need that selfbot VoiceConnection CONNECTED.
Fix: establish the selfbot VoiceConnection eagerly in recorder.startRecording,
BEFORE joinVoiceChannel, so it rides the bot's fresh join (Discord emits
VOICE_SERVER_UPDATE → selfbot authenticates). videoRecorder reuses the cached
connection per guild, tears it down on voice stop/destroy. Best-effort — never
blocks audio recording.
Verified: typecheck + build + biome (src/) green; 9/9 videoRecorder tests.
Root cause of why Phase A/B captured zero video: @discordjs/voice is audio-only
and never sends the gateway STREAM_WATCH signal, so Discord never forwards a
member's video RTP to the bot. Live diagnostic confirmed: while members were
sharing, audio .ogg files flowed for many users but no non-opus RTP ever
arrived.
Correct path: discord.js-selfbot-v13 ships a complete native watch/record stack.
New src/modules/voice-recording/videoRecorder.ts:
- Detects streamers via voiceState.streaming on a single idempotent
voiceStateUpdate listener.
- client.voice.joinChannel() (reuses the single session alongside
@discordjs/voice) + joinStreamConnection(userId) -> STREAM_WATCH (op 20).
- receiver.createVideoStream(userId, path) -> PacketHandler -> Recorder
(ffmpeg over UDP loopback) -> Matroska .mkv, decryption handled internally.
Wired in recorder.ts (trackChannel/untrackChannel) + bootstrap.ts
(setVideoRecorderClient/setVideoRecordingsDir). All best-effort; failures log
and never break existing voice/audio. Unit tests 6/6 (single listener, watch
handshake + mkv path, skip own video, idempotence, teardown). Full suite
20 files / 176 tests green; typecheck + build + biome clean.
UNVERIFIED live yet: needs deploy + a streamer to confirm Recorder ready +
playable .mkv.
Temporary diagnostic to answer definitively whether Discord actually delivers
video RTP to the bot when someone screen-shares / turns on camera. Both
videoReceiver and screenShareAudio rely on ssrcMap emitting videoSSRC from
voice-state updates, and NO "Video SSRC appeared"/"Screen-share video started"
lines appear even after a real share+record. This logs any RTP packet whose
payload type is not Opus (120) so we can tell: (a) video RTP IS arriving but
attribution/signaling fails, vs (b) Discord sends no video at all to a
non-signaling receiver. Remove this log once the gap is understood.
ssrcMap emits "delete" when a VoiceUserData is removed (user stops sharing /
leaves voice). Hook it to close the user's open video burst ~500ms later so the
ffmpeg mux starts as soon as they stop, instead of waiting up to 5s for the idle
sweep. No-op if no burst exists; safe on normal teardown.
Phase A captured raw .h264 streams but left them as non-playable elementary
streams. Phase B adds automatic muxing: when a video burst closes, the raw
.h264 is remuxed to a self-contained MP4 via `ffmpeg -c copy` (no re-encode,
fast) with `+faststart`, waits for the write stream to fully flush first so the
mux never reads a truncated tail, and deletes the raw .h264 on success (keeping
it on failure). Output: <RECORDINGS_DIR>/<uid>/video-<ssrc>-<ts>.mp4.
muxToMp4 is exported + covered by a real-ffmpeg vitest (tests/videoReceiver.test.ts):
generates a tiny baseline h264, remuxes, asserts mp4 exists/non-empty & raw deleted
(also the 5 depacketizer tests). Full gateway suite 170/170 green, tsc + biome clean.
@discordjs/voice only decrypts/forwards AUDIO (opus) — its onUdpMessage drops
every non-opus RTP packet (dist/index.mjs:2068 `!== RTP_OPUS_PAYLOAD_TYPE`, 120).
Video RTP (H264 camera + screen share, plus VP8/VP9/AV1) arrives on the same UDP
socket but was silently discarded.
New videoReceiver.ts wraps receiver.onUdpMessage (like screenShareAudio.ts):
- detects video payload types (96/98/101/102/106/116/126/127),
- decrypts them with the connection secret key/encryptionMode via the SAME
receiver.parsePacket path @discordjs/voice uses for audio (so DAVE + voice
encryption are handled identically),
- depacketizes H264 to AnnexB (single NAL, STAP-A, and FU-A fragmentation),
waiting for a keyframe (SPS/PPS/IDR) before writing,
- writes a raw .h264 elementary stream per user per burst under
<RECORDINGS_DIR>/<uid>/video-<ssrc>-<ts>.h264.
Attribution: a videoSSRC→user index is built from ssrcMap updates; a proximity
fallback mirrors screenShareAudio's inferScreenShareOwner. Bot's own video is
skipped.
Unit tests: tests/videoReceiver.test.ts (AnnexB start code, keyframe gating,
FU-A reassembly, orphan-fragment tolerance) — 5/5 green.
Phase A only (capture raw h264). Phase B (ffmpeg decode+mux to MP4/WebM +
persist) and Phase C (frontend playback) are follow-ups.
A server-muted/server-deafened bot can't reliably receive/record members' audio
(and definitely can't receive video/screen share). After the voice connection
is Ready, force a REST guild-members PATCH (mute:false, deaf:false) on the self
member so the bot is auto-unmuted & undeafened on every join/reconnect.
Requires MUTE_MEMBERS + DEAFEN_MEMBERS permissions (user granted). Failure is
logged as a warning and never breaks the voice join.
Previously isAgeRestrictedMessage() early-returned in messageCreate/messageUpdate,
so NSFW-channel messages were never stored at all. Now they are captured like
any message (visible in dashboard), while the existing age-restricted skip path
(queueMessageAnalysis -> buildAgeRestrictedSkipResult) marks them clean with flag
age_restricted WITHOUT calling the LLM.
NSFW content is also deliberately kept OUT of the Qdrant public semantic-search
archive (archiveMessageEmbedded skips when isAgeRestricted), so it can't be found
via public web search. No schema change needed (metadata already carries channel.nsfw).
Previously isAgeRestrictedMessage() early-returned in messageCreate/messageUpdate,
so NSFW-channel messages were never stored at all. Now they are captured like
any message (visible in dashboard), while the existing age-restricted skip path
(queueMessageAnalysis -> buildAgeRestrictedSkipResult) marks them clean with flag
age_restricted WITHOUT calling the LLM.
NSFW content is also deliberately kept OUT of the Qdrant public semantic-search
archive (archiveMessageEmbedded skips when isAgeRestricted), so it can't be found
via public web search. No schema change needed (metadata already carries channel.nsfw).
Root-cause fixes for 'banyak miss & terpotong' in the voice->recording flow:
- subscribe BEFORE collecting user metadata. receiver.speaking 'start' fires
on the FIRST opus packet, and onUdpMessage forwards frames to the
subscription only when one exists — every frame during the old
await collectUserMetadata (a Discord REST roundtrip on cache miss) was
dropped, cutting off the start of every burst. Now subscribe synchronously
(guard first, no await in between), then fetch metadata in the background
and discard the burst if the speaker turns out to be a bot.
- one segment per burst: drop the fixed 5s RECORDING_SEGMENT_MS rotation on
the OGG path, which split continuous speech mid-word/sentence. Only the
web-PCM decoder still rotates (bounds memory).
- finalize only once the underlying file has flushed to disk (wait on the
write stream 'finish'), so upload/transcode reads a complete file.
- raise AfterSilence 3000->4000ms so natural pauses (thinking, interruptions)
don't split one utterance into several recordings.
- lower the 'too short to keep' threshold 1000->300ms so brief replies
("ya", "siap") are kept instead of dropped.
All typecheck / biome(src/) / vitest (164) green.
- nativeBuildInputs: remove cmake, rustc, cargo, git — GMW's only native deps (@discordjs/opus, sharp) are PREBUILT, no source compile needed (cmake/rust were inherited for node-datachannel which is 9router, not GMW). python3/gnumake/gcc stay as node-gyp fallback for opus.
- pruneProd: also strip .pnpm/@types+* (pure TS decls pulled into the prod graph as real deps by discord-api-types/pg-protocol, never required at runtime) and .pnpm/opusscript@* (pure-JS fallback Opus engine that prism-media only loads IF native @discordjs/opus fails — native is always present, so opusscript is never executed).
- Verified: nix flake check OK; typecheck + biome check src/ green; runtime smoke test post-prune loads @discordjs/voice, sharp, selfbot, tiktoken, piscina and encodes a frame via native opus.
- mediaDownloader: flip URL candidate order so discord_url is tried
before uploaded_url (uploaded_url is archive-only fallback)
- ai-analysis-worker: remove upload-pending race guard that blocked
analysis until Tele upload completed; analysis now runs immediately
on the Discord CDN URL
- batchProcessor: remove upload-pending defer/poll-backoff logic
- individualFallbackProcessor: remove upload_pending requeue loop
- batchOutcomeClassifier/fallbackResultClassifier: drop upload_pending
classification (no longer needed)
- tests: update batchOutcomeClassifier + fallbackResultClassifier tests
to reflect removed upload_pending signal
Switch GMW's AI LLM base URL from 9router (https://9router.asepharyana.my.id/v1)
to omniroute on imrnes (http://100.121.180.82:20128/api/v1).
- Update default AI_LLM_BASE_URL in discord-gateway + backend config schemas
- Update .env.example documentation
- Update all 9router references in comments/docs/tests to omniroute
- Production BWS secret gmw_ai_llm_base_url already updated
Omniroute uses /api/v1 prefix (not /v1 like 9router), so the base URL
now correctly points at the right API path for the OpenAI SDK.
Root cause of 'navbar mobile tak bisa pindah halaman': Next <Link>
client-side navigation is dead app-wide. A React hydration mismatch
(#418: 'server rendered text didn't match the client') is thrown by the
SSR-seeded live feeds — relative times (formatRelativeTime(e.edited_at) /
m.created_at) computed with Date.now() render slightly differently on
server vs client, which breaks the Next client router (router.push is a
no-op). The desktop NavRail worked only because it uses plain <a href>
(hard navigation bypasses the broken router).
Fixes:
- mobile-nav.tsx: use plain <a href> (NOT Next <Link>), identical to the
working sidebar NavRail, so mobile nav always navigates regardless of
router state ('ikuti cara kerja sidebar').
- Add suppressHydrationWarning to the SSR-seeded relative-time spans so
server/client drift no longer throws #418 (EditHistory, LiveModerationFeed,
messages/results + detail rows, recordings, TermGlossary,
ChannelCultureGlossary, CategoryDrilldown).
Verified on non-prod :4024 @375px: Voice/Media/Search all navigate, no #418
in console. Plan: .hermes/plans/mobile-nav-hydration-fix.md
Use fixed w-[72px] shrink-0 pills (not flex-1) so the 7 bottom-nav items
keep their shape and scroll horizontally instead of compressing labels to
overlap; add truncating 10px mobile label (12px >=sm).
- Root cause of 'navbar mobile tak bisa pindah halaman': the chatbot FAB
(fixed right-4 bottom-5 z-50) overlapped the rightmost mobile bottom-nav
items (z-40) and intercepted taps. Raise the FAB above the nav on mobile
(bottom above nav, md:bottom-5 on desktop) so it never blocks nav taps.
- Mobile bottom nav now mirrors the FULL desktop sidebar (all 7 items:
dashboard, messages, voice, media, recordings, moderation, analysis);
horizontally scrollable + snap-to-active on narrow screens.
- Select dropdown: clamp portal position within viewport + min-width so it
never overflows off-screen on mobile triggers near the right edge; larger
tap targets on touch.
- Input/Textarea: text-base (16px) on touch to prevent iOS auto-zoom on
focus, text-sm on >=sm; comfortable mobile min-height for chat input.
- Add .hermes/plans/mobile-nav-form-responsive.md spec.
getRecentEdits SELECT ... m.content AS new_content returned the message's
ORIGINAL content (never updated on edit) instead of the post-edit content.
messages.content stores the original body; the current/last-edited body lives
in messages.edited_content. So before===after in the Message Edits diff.
Fix: COALESCE(m.edited_content, m.content) AS new_content so 'After' shows the
edited text and diffs against the captured before-content are meaningful.
Prevent recurring infinite restart loop (389x crash) caused by drizzle
re-applying already-applied migrations when public.__drizzle_migrations
tracking is empty/partial.
- 0017/0018: ADD COLUMN IF NOT EXISTS (re-run safe)
- 0019: DO-block rename that handles all prior states (server_name-only,
both columns, or server_nick-only) so it never errors or double-renames
- seedDrizzleHistory: reconcile tracked created_at to the journal's latest
'when' when the schema already reflects the latest migration, instead of
early-returning on an existing-but-empty/partial tracking table
Extract server_nick from metadata.member.displayName in backend messageMapper,
add server_nick to frontend MessageRecord type, and update messages view +
analysis view to display the member's server-specific nickname (with @username
as secondary context) instead of the global username.
Rename server_name (guild name) to server_nick and populate it from
the member's server-specific display name (metadata.member.displayName)
at write time. This is what the moderation dashboard should show as
TARGET — e.g. server nick 'Bandar Togel「✔ ᵛᵉʳᶦᶠᶦᵉᵈ 」' for global
username '.nichiyobi'. Backfilled 210 existing actions from messages
metadata (reset_nickname rows now show 'Sarjana .jav', 'Penindas
Minoritas', etc). Frontend TARGET shows server nick with global
username as secondary context.
Replace static OFFENSIVE_USERNAME_KEYWORDS substring matching with a
lightweight LLM call that evaluates whether a global username violates
server rules (gambling, scam, NSFW, SARA, etc). Fail-open design:
if the LLM call fails/times out, the nickname reset still completes.
After resetting an offensive server nickname to the global username,
check the global username against gambling/scam keyword list. If it
also violates, generate a random 'UserXXXXX' nickname to prevent
circumvention via offensive global usernames.
Denormalize guild name alongside username so the moderation dashboard
shows both TARGET and server even after message table purges.
Migration 0018. Frontend displays 'username · server_name' in TARGET.
Gateway (transmitter.ts):
- Auto-stop on FFmpeg crash: non-zero exit triggers stop() to prevent
silent audio loss and resource leaks
- Voice activity timeout (10s): auto-stops transmitter when no PCM
received, preventing dead-air CPU waste on backgrounded tabs
- Stderr cap (4KB): prevents unbounded memory growth in long sessions
Gateway (voice.handler.ts):
- Double-check voiceController.getStatus().connected before starting
transmitter — detects stale player state after gateway disconnect
Frontend (context.tsx):
- Force-refetch voice status on WS reconnect — UI converges in <1s
instead of waiting up to 4s for SWR poll interval
All: tsc clean, biome clean
- Updated RootLayout to include a noise overlay class.
- Refactored LiveModerationFeed to use CSS animations for item entry.
- Simplified AmbientCanvas by removing WebGL and using pure CSS for ambient effects.
- Converted Chatbot to use CSS for floating window animations.
- Updated Card component to include a gradient border.
- Refactored PageTransition to use CSS for fade-scale animations.
- Enhanced SectionHeader with reveal-up animation class.
- Replaced GSAP animations in NavRail with CSS animations for item entry.
- Refactored VoiceStage to use CSS for speaker node animations and pulse rings.
- Removed GSAP dependencies from use-gsap-animation hook, implementing CSS-based stagger reveal.
- getMessageById: run findById and getEditHistory in parallel via
Promise.all instead of sequential awaits (halves latency for
message detail view)
- Drop 3 unused indexes on messages table: idx_messages_guild_ai_status_created
(0 scans), idx_messages_guild_ai_status_analyzed (97 scans),
idx_messages_guild_created_deleted (95 scans, superseded by covering index)
— saves ~5.9MB index space + reduces write amplification
- Add covering index idx_messages_guild_created_covering for the primary
findMany query pattern (guild_id + created_at DESC with INCLUDE columns)
Two bugs causing 23 spurious 'error' logs after successful deletions:
1. batchProcessor switch missing 'completed' case: partitionBatchOutcome
returns 'completed' for successful messages, but the switch only handled
'upload_pending' and 'api_failed'. Successful messages fell through to
default → re-enqueued to individual fallback → re-analyzed → re-delete
attempt → error (message already gone from Discord). Now explicitly
skips 'completed' messages.
2. isAlreadyDeletedError only caught codes 10008/404. Discord also returns
10003 (Unknown Channel) and 50001 (Missing Access) when a message or
channel is gone. Added these codes plus text-based fallback matching
'Unknown Message'/'Unknown Channel'.
Impact: eliminates ~23 redundant error logs per day + stops wasted LLM
calls re-analyzing already-processed messages.
isEligibleForAutoDelete was rejecting messages where recommendedAction
was 'warn' or 'review' — only 'delete' and 'escalate' were accepted.
This caused 28+ medium-severity flagged messages to be logged as
'not_eligible' instead of being auto-deleted.
Changes:
- deriveRecommendedAction: return 'delete' for flagged+medium severity
(previously only critical/high triggered delete; medium got 'review')
- isEligibleForAutoDelete: accept 'warn' as valid recommendedAction
alongside 'delete' and 'escalate'
Impact: ~28 pending medium-severity flagged messages + all future
'warn'-action flagged messages will now be eligible for auto-deletion.
- Refactored ChannelCultureGlossary, LiveModerationFeed, TermGlossary, and Chatbot components to use new design tokens and styles.
- Updated button, badge, and section components to align with the new design system.
- Enhanced the visual hierarchy and accessibility of various UI elements.
- Improved responsiveness and hover states across components.
- Replaced hardcoded colors with design tokens for better maintainability.
- Cleaned up imports and removed unused hooks in ModerationView and RecordingsView.
- Updated UI elements for better visual consistency and accessibility.
- Improved performance by optimizing rendering logic and reducing unnecessary state updates.
- Added filtering functionality in ChannelCultureGlossary and TermGlossary components.
- Enhanced LiveModerationFeed with GSAP animations for smoother transitions.
- Updated chatbot component for better user experience and responsiveness.
- Refined NavRail component for cleaner code and improved navigation item rendering.
- Switch AI_LLM_BASE_URL from omniroute.imrnes.team to 9router.asepharyana.my.id
- Keep AI_LLM_MODEL as 'text' (9router uses alias-based routing, not bare names)
- Update .env.example comments to document 9router
- Per user: multimodal stays 'multimodal' alias
API verified: curl to 9router/v1/chat/completions with model 'text'
returns HTTP 200 (OpenAI-compatible format)
- Switch AI_LLM_BASE_URL from omniroute.imrnes.team to 9router.asepharyana.my.id
- Keep AI_LLM_MODEL as 'text' (9router uses alias-based routing, not bare names)
- Update .env.example comments to document 9router
- Per user: multimodal stays 'multimodal' alias
API verified: curl to 9router/v1/chat/completions with model 'text'
returns HTTP 200 (OpenAI-compatible format)
- Change AI_LLM_BASE_URL default from omniroute.imrnes.team to 9router.asepharyana.my.id
- Update AI_LLM_MODEL default from 'text' to 'claude-opus-5' (bare model name
compatible with 9router/OpenAI-compatible router)
- Update .env.example and inline comments to reflect 9router
- discord-gateway config now matches backend (which already uses 9router)
Batch race guard balikin {ok:true, rows:[]} tanpa sinyal saat semua target
masih upload-pending -> processor klasifikasi semua incomplete -> fanout ke
individual queue -> di situ requeue + reschedule 250ms -> balik ke batch:
hot loop ~300ms sepanjang upload (10 siklus/3 dtk di log prod 08:13).
Fix: worker batch kini return uploadPendingIds eksplisit; classifier pure
baru (partitionBatchOutcome) partisi completed/upload_pending/incomplete/
parse_failed/api_failed; target upload-pending DEFERRED dengan poll backoff
linear (AI_ANALYSIS_UPLOAD_POLL_MS 1500 base, cap AI_ANALYSIS_MAX_UPLOAD_POLL_MS
8000), tidak pernah masuk fanout; tail shouldScheduleNext tak menimpa defer.
Test: tests/batchOutcomeClassifier.test.ts (8 kasus, pure tanpa DB/Piscina).
Analisis pertama tetap berkonteks (chat history) demi akurasi, tapi
verdict clean non-actionable (conf>=0.85) juga ditulis di bare key
tanpa konteks. Repeat teks sama di channel lain -> exact cache HIT,
bukan LLM call baru. Guard sama dgn read path; dedupe LRU per proses;
bare row tanpa embedding (tier semantic sudah global).
- Fase-1 exact-cache lookup: N query serial -> SATU query ANY($1::text[])
- Global reuse utk bare key legacy, HANYA verdict non-actionable
(clean/flagless/action=none, conf>=0.85, umur<=72h) — flagged/warn
tetap context-scoped
- Semantic cache dua-band: clean band 0.92 default, actionable tetap
0.97; di antara band -> LLM (fail-open ke akurasi)
- hit_count kini di-increment (bulk UPDATE per batch) -> hit-rate terukur
- Cache hasil wikipediaSearch di Redis (6h, hanya hasil non-kosong)
- Memoize fetchUrlSafely utk type=text (LRU 30m + in-flight dedupe)
- makeImageCacheKey strip query CDN Discord (?ex/is/hm, format/width)
-> attachment sama = satu key vision, skip re-download+re-vision
Spec: .hermes/plans/2026-08-24-ai-analysis-cache-optimization.md
Tests: +33 (cacheGuards, discordImageKeyNormalize, cacheBatchLookup)
- pickBatchWithinBudget: stop di overflow pertama (break), bukan skip —
batch tetap prefix kronologis tanpa gap analisis di tengah timeline.
Diekstrak ke batchBudget.ts (pure, estimator di-inject) + regression test.
- callModerationLLM: param opsional maxTokens; text/media caller menghitung
ceiling dari estimasi prompt (floor 2048, cap 16384) — batch kecil tak
lagi reserve window completion 16k.
- getPending/IncompleteMessagesByConversation: sort hasil UPDATE..RETURNING
by created_at ASC — Postgres tak menjamin urutan, konsumen (anchor konteks
messages[0], prefix batch) bergantung pada urutan kronologis.
- normalizeStoredStatus(): exact-hash & semantic (Qdrant/PG) cache reader
sebelumnya menipiskan 'warn' jadi 'flagged'/'clean' (type narrowing
legacy clean|flagged) — merusak gating auto-delete & label dashboard.
Kini status tersimpan dipertahankan penuh (clean/warn/flagged).
- prompts: hapus referensi <user_history> yang tak pernah di-inject,
SearXNG -> Wikipedia (sudah migrasi), referensi section yang tak ada,
typo 'secifik', dan baris list rusak '|-'.
- moderationBuilders: buang dead code buildUserProfilesBlock/
buildUserProfileRef/UserProfileEntry/buildUserHistoryXml (tanpa caller
produksi sejak context minimization) + test-nya.
- test baru: tests/storedStatusNormalization.test.ts (regresi warn).
upload.asepharyana.my.id redirect ke file mp3 tunggal; generic extractor
yt-dlp expose format ID '0' sehingga '-f bestaudio' gagal 'Requested
format is not available'. Chain bestaudio[ext=m4a]/bestaudio/best tetap
dapat m4a di YouTube dan jatuh ke 'best' untuk direct file.
- Recordings: custom RecordingAudioPlayer (play/pause, buffering spinner,
click-to-seek, time label, eq bars, single-playback antar kartu) +
highlight kartu now-playing
- Media: thumbnail di disc hero + queue row, equalizer saat playing,
badge 'up next', label Paused vs Now playing
- MiniPlayer global di AppFrame (fixed bottom-right, hidden on /media)
menggantikan use-media-player.tsx dead provider (dihapus)
- Voice: mic level meter live (AnalyserNode RMS) + slider mic/listen volume
tsc reported 'Property some does not exist on type {}' on doc.tags
because Drizzle jsonb inference returns a generic object. Cast explicitly.
This unblocks the GMW GitHub Actions deploy (test job).
The Drizzle schema used pgText('tags').array() which emits a Postgres
text[] column, but the migration defines tags as jsonb. The mismatch caused
INSERT/LIST on materi_documents to throw INTERNAL_SERVER_ERROR (500) because
Drizzle sent a text[] where the column expected jsonb.
- schema.ts: tags → pgJsonb('tags').notNull().default('[]')
- migration: dropped+recreated to match schema (jsonb, ms epoch defaults,
owner_user_id default 'anonymous')
Two independent WS handlers (useMessagesWsSync for message_created/
updated/analyzed, and useMessagesStream for message_snapshot) both
prepend live messages to the SWR list without enforcing order.
When frames arrive out-of-order (common with batched WS delivery),
the message feed gets scrambled.
Fix: add sortMessages() helper that sorts newest-first by created_at
(the list's stored order before .reverse() for display) and apply it
in every patchLists/mutate updater: message_created, message_updated,
message_analyzed, message_snapshot, and useLoadMore page appends.
Function declaration is hoisted so useLoadMore (defined above the
helper) can use it.
The fix-imports.mjs script blindly appended '.js' to every @/ alias
import, even when the source specifier already carried a .js
extension (e.g. '@/shared/config/index.js'). This produced
'index.js.js' in the emitted dist/, causing ERR_MODULE_NOT_FOUND
at startup.
This was latent: only triggered once digestScheduler.ts (which
uses @/shared/config/index.js with explicit extension) was built.
The user-reputation removal (2a8f6d9) was also blocked by this
bug — stale binary kept crashing with 'user_reputations' query
errors because it was never redeployed.
Fix: only append .js when the @/ specifier has no existing
extension. Applied to both gateway and backend scripts.
SearXNG was already replaced by Wikipedia REST/Action APIs (wikipediaClient.ts).
Update comments to reflect the current implementation: term glossary now
resolves definitions via Wikipedia → Redis → Postgres cache chain, with no
SearXNG dependency.
Discord GoLive sends screen-share audio on a separate SSRC from the
user's microphone. In @discordjs/voice v0.19, VoiceReceiver.onUdpMessage
silently drops packets for SSRCs not in ssrcMap (which is only populated
from VOICE_STATE_UPDATE/VOICE_SERVER_UPDATE). This caused screen-share
audio to never trigger receiver.speaking and never reach the speakingHandler.
Fix: hookScreenShareAudio() wraps onUdpMessage to:
1. Detect incoming RTP packets with unknown SSRCs (OPRUS payload type 120)
2. Infer the owning userId by proximity to known audioSSRC
3. Clone the user's VoiceUserData into ssrcMap under the new SSRC
4. Let the original handler decrypt and forward to the subscription stream
5. Listen on ssrcMap 'create'/'update' events for video SSRC changes
Also removes the broken initial approach (polling ssrcMap which never
contains screen-share SSRCs).
Automated public weekly summary: top categories/domains/channels + coverage rate, posted to configured webhook. Uses getDatabase() direct query (no oRPC HTTP dependency), guards one-fire-per-week on restart.
Gateway @/ alias imports use no .js extension (relative imports
keep .js). The .js suffix on @/ paths caused double-extension
ERR_MODULE_NOT_FOUND (embeddingClient.js.js) at runtime.
GMW (Guild Moderation Watcher) is a Discord bot + web dashboard for AI-powered moderation. A monorepo with three services: a selfbot gateway that captures Discord events and runs LLM moderation, an Express/oRPC backend that serves the dashboard API, and a Next.js 16 SSR frontend. They communicate via Redis pub/sub (gateway→backend) and WebSocket (backend→browser).
## Quick reference
```bash
# Per-service — cd into the service first
pnpm install # install deps (pnpm 11, not npm or bun)
pnpm typecheck # tsc --noEmit
pnpm lint # biome check
pnpm format # biome format --write
pnpm build # gateway/backend: tsc + fix-imports.mjs; frontend: next build
pnpm test# vitest run (gateway & backend only — frontend has no tests)
```
No monorepo-level scripts exist. Run each command from inside the service directory.
## Layout
```
services/
├── discord-gateway/ Event-driven selfbot. No HTTP (except :4016 metrics).
│ ├── src/
│ │ ├── app/ Bootstrap, shutdown, retention
│ │ ├── modules/ Feature modules — each self-contained
│ │ └── shared/ Config, DB (Drizzle), logger, errors, utils
│ ├── tests/ Vitest tests (colocated, not inside src/)
- **pnpm** (v11), not npm or bun. Lockfiles are committed. Node ≥ 22.
- ESM throughout (`"type": "module"` in all package.json files).
### Import style
Source files use `@/*` path aliases (mapped in tsconfig to `./src/*`). Relative imports **must include the `.js` extension** (e.g., `from "./embed.js"`). The `moduleResolution: "bundler"` tsconfig setting allows bare specifiers during dev, but `tsc` emits them as-is. A post-build script (`scripts/fix-imports.mjs`) rewrites both `@/` aliases and extensionless imports in `dist/` so Node ESM can resolve them at runtime.
### Error handling
Both gateway and backend define an `AppError` base class in `@/shared/errors/index` with subclasses: `ValidationError` (400), `NotFoundError` (404), `UnauthorizedError` (401), `DatabaseError` (500), `ConfigError` (500). Services throw these; callers or middleware map them to HTTP status codes.
### Logging
Use `createChildLogger('module-name')` from `@/shared/logger/index`. Never use raw `console`. It wraps pino; in development it pretty-prints via `pino-pretty`.
### Config
Environment variables are validated with Zod at startup in `shared/config/index.ts` of both gateway and backend. Do not read `process.env` directly outside config modules.
### Module boundaries
- Gateway: each feature lives in `src/modules/<name>/` with its own `index.ts` barrel. Modules register event listeners and are composed in `src/app/bootstrap.ts`.
- Backend: `modules/<name>/` follows schema → repository → service → controller → routes. Data flows up only. No cross-module repository imports.
- Frontend: `page.tsx` is a server component that fetches data via `src/lib/api/server.ts` (oRPC over HTTP, server-side only) and passes it to a `view.tsx` client component. Browser code uses `src/lib/orpc/client.ts` (oRPC over WebSocket via partysocket). Never import the server API client from a client component.
### API layer
The backend exposes an oRPC router mounted at `/trpc` (both HTTP and WebSocket). The frontend does **not** use a REST `/api/*` layer — all data goes through oRPC procedures. The server-side client (`src/lib/api/server.ts`) uses a fetch-based RPCLink; the browser client uses a WebSocket-based RPCLink backed by partysocket for auto-reconnection. Results are asserted to the frontend's local types at each call site (the backend's router type is not imported into the frontend).
### Testing
- **Gateway**: Vitest, tests in `tests/` at the service root. Config sets env vars (`DISCORD_TOKEN`, `DATABASE_URL`, etc.) so tests run without real services.
- **Backend**: Vitest, tests in `tests/` and `src/`. Includes an `e2e.test.ts` excluded from CI (needs a live backend).
- **Frontend**: No test runner configured.
- Tests are pure-function / unit-level. Mock external dependencies; do not start real DB/Redis in tests.
1.**Never commit without running `fix-imports.mjs` after `tsc`** — gateway and backend builds will produce ESM that crashes at startup (`ERR_MODULE_NOT_FOUND`).
2.**Don't add `@discordjs/opus` build-from-source flags** — it ships prebuilt binaries for Node 22. Forcing source builds in CI/Nix will fail or add minutes of compile time.
3.**Frontend SSR fetches are always `cache: "no-store"`** — the server API client bypasses Next.js fetch cache. Do not add caching without understanding the live dashboard requirement.
4.**oRPC types are loosely coupled** — the frontend casts oRPC results to its own types with `as unknown`. Adding a field to the backend schema does not automatically update the frontend type. Update both sides.
5.**Gateway is a selfbot** (`discord.js-selfbot-v13`) — it uses a user token, not a bot token. It must not be deployed as a standard Discord bot.
description:"Aeronet Visualization Dashboard Section is designed for demonstrating application workflows and interface hierarchy. Key features include clear information density, modular panels, and interface rhythm. It is suitable for product showcases, admin panels, and analytics experiences."
colors:
primary:"#FFFFFF"
secondary:"#000000"
accent:"#FFFFFF"
background:"#FFFFFF"
surface:"#18181B"
text-primary:"#111827"
text-secondary:"#4B5563"
border:"#E5E7EB"
typography:
display-lg:
fontFamily:"Inter"
fontSize:"64px"
fontWeight:500
lineHeight:"1.04"
letterSpacing:"0"
body-md:
fontFamily:"Inter"
fontSize:"16px"
fontWeight:400
lineHeight:"1.6"
label-md:
fontFamily:"JetBrains Mono"
fontSize:"12px"
fontWeight:600
lineHeight:"1.2"
spacing:
base:"8px"
gap:"16px"
card-padding:"24px"
section-padding:"80px"
rounded:
card:"16px"
control:"8px"
pill:"9999px"
components:
card:
background:"Use the surface token with subtle borders and HTML-matched shadow depth"
radius:"Match the declared card radius token"
button:
background:"Use primary or accent colors for the main action"
radius:"Use the control or pill radius based on the source HTML"
---
# AeroNet Visualization
Source: Neuform Featured templates from top creators. Author: Sourasith Phomhome (@madebysourasith). Views: 419; favorites: 14; remixes: 4.
Aeronet Visualization Dashboard Section is designed for demonstrating application workflows and interface hierarchy. Key features include clear information density, modular panels, and interface rhythm. It is suitable for product showcases, admin panels, and analytics experiences.
Use the attached HTML reference as the source of truth. Preserve the visible hierarchy, first-screen composition, section rhythm, density, and interaction tone before adapting copy or content.
Key visible headings include: Planet Core.
## Colors
Anchor the palette in primary #FFFFFF, secondary #000000, accent #FFFFFF, background #FFFFFF, surface #18181B, text-primary #111827. Keep background, surface, text, and border roles distinct so generated layouts retain the same contrast pattern as the source.
## Typography
Use Inter for display moments and Inter for body copy unless the HTML clearly demands a compatible fallback. Labels and technical metadata should use JetBrains Mono or an equivalent mono face.
## Layout
Keep spacing deliberate and stable. Favor the same grid direction, max-width behavior, card density, and responsive stacking seen in the HTML. Do not replace distinctive source structures with generic SaaS sections.
## Components
Dashboard, chart, and data panels should preserve their compact operational hierarchy, nested surfaces, and metric emphasis.
## Motion
Preserve existing motion cues such as masked reveals, staggered entrance, hover lift, scroll-triggered transitions, and ambient movement. Keep easing smooth and restrained.
## WebGL & Effects
If the source includes canvas, WebGL, Three.js, gradients, particles, or atmospheric effects, rebuild them as supporting layers behind the content. Keep effects performant, responsive, and secondary to the interface.
## Guardrails
- Do not flatten the source into a generic card grid.
- Do not swap the color mode unless the source clearly supports it.
- Preserve the first viewport signal, focal object, and visual density.
- Keep buttons, cards, and badges aligned to the same radius and border language.
description:"Nexus Capital Feature Section is designed for highlighting product capabilities and value points. Key features include reusable structure, responsive behavior, and production-ready presentation. It is suitable for component libraries and responsive product interfaces."
colors:
primary:"#F2EAD3"
secondary:"#000000"
accent:"#F2EAD3"
background:"#000000"
surface:"#F2EAD3"
text-primary:"#FFFFFF"
text-secondary:"#A1A1AA"
border:"#27272A"
typography:
display-lg:
fontFamily:"Inter"
fontSize:"64px"
fontWeight:500
lineHeight:"1.04"
letterSpacing:"0"
body-md:
fontFamily:"Playfair Display"
fontSize:"16px"
fontWeight:400
lineHeight:"1.6"
label-md:
fontFamily:"JetBrains Mono"
fontSize:"12px"
fontWeight:600
lineHeight:"1.2"
spacing:
base:"8px"
gap:"16px"
card-padding:"24px"
section-padding:"80px"
rounded:
card:"16px"
control:"8px"
pill:"9999px"
components:
card:
background:"Use the surface token with subtle borders and HTML-matched shadow depth"
radius:"Match the declared card radius token"
button:
background:"Use primary or accent colors for the main action"
radius:"Use the control or pill radius based on the source HTML"
---
# Nexus Capital Render
Source: Neuform Featured templates from top creators. Author: Meng To (@mengto). Views: 311; favorites: 22; remixes: 15.
Tags: feature, section, animated, bento, charts.
## Overview
Nexus Capital Feature Section is designed for highlighting product capabilities and value points. Key features include reusable structure, responsive behavior, and production-ready presentation. It is suitable for component libraries and responsive product interfaces.
NEXUS WEALTH CORP. Beyond traditional equity. The multi-asset strategy. A dynamic portfolio framework where capital transitions seamlessly from baseline reserves to high-yield synthetic instruments. Drag the timeline to…
## Composition
Use the attached HTML reference as the source of truth. Preserve the visible hierarchy, first-screen composition, section rhythm, density, and interaction tone before adapting copy or content.
Key visible headings include: Beyond traditional equity. The multi-asset strategy..
## Colors
Anchor the palette in primary #F2EAD3, secondary #000000, accent #F2EAD3, background #000000, surface #F2EAD3, text-primary #FFFFFF. Keep background, surface, text, and border roles distinct so generated layouts retain the same contrast pattern as the source.
## Typography
Use Inter for display moments and Playfair Display for body copy unless the HTML clearly demands a compatible fallback. Labels and technical metadata should use JetBrains Mono or an equivalent mono face.
## Layout
Keep spacing deliberate and stable. Favor the same grid direction, max-width behavior, card density, and responsive stacking seen in the HTML. Do not replace distinctive source structures with generic SaaS sections.
## Components
Dashboard, chart, and data panels should preserve their compact operational hierarchy, nested surfaces, and metric emphasis.
## Motion
Preserve existing motion cues such as masked reveals, staggered entrance, hover lift, scroll-triggered transitions, and ambient movement. Keep easing smooth and restrained.
## WebGL & Effects
If the source includes canvas, WebGL, Three.js, gradients, particles, or atmospheric effects, rebuild them as supporting layers behind the content. Keep effects performant, responsive, and secondary to the interface.
## Guardrails
- Do not flatten the source into a generic card grid.
- Do not swap the color mode unless the source clearly supports it.
- Preserve the first viewport signal, focal object, and visual density.
- Keep buttons, cards, and badges aligned to the same radius and border language.
description:"Summit Cloud Pricing Section is designed for comparing plans and supporting conversion decisions. Key features include plan comparison blocks and conversion-oriented actions. It is suitable for subscription pricing pages and plan comparison experiences."
colors:
primary:"#2563EB"
secondary:"#050510"
accent:"#3B82F6"
background:"#050510"
surface:"#18181B"
text-primary:"#FFFFFF"
text-secondary:"#A1A1AA"
border:"#27272A"
typography:
display-lg:
fontFamily:"Inter"
fontSize:"64px"
fontWeight:500
lineHeight:"1.04"
letterSpacing:"0"
body-md:
fontFamily:"Inter"
fontSize:"16px"
fontWeight:400
lineHeight:"1.6"
label-md:
fontFamily:"JetBrains Mono"
fontSize:"12px"
fontWeight:600
lineHeight:"1.2"
spacing:
base:"8px"
gap:"16px"
card-padding:"24px"
section-padding:"80px"
rounded:
card:"23px"
control:"18px"
pill:"9999px"
components:
card:
background:"Use the surface token with subtle borders and HTML-matched shadow depth"
radius:"Match the declared card radius token"
button:
background:"Use primary or accent colors for the main action"
radius:"Use the control or pill radius based on the source HTML"
Summit Cloud Pricing Section is designed for comparing plans and supporting conversion decisions. Key features include plan comparison blocks and conversion-oriented actions. It is suitable for subscription pricing pages and plan comparison experiences.
Summit Platform Solutions Pricing Docs Start Migration Now in General Availability Migrate to the cloud with confidence Automated discovery, dependency mapping, and zero-downtime migration for enterprise workloads. Move…
## Composition
Use the attached HTML reference as the source of truth. Preserve the visible hierarchy, first-screen composition, section rhythm, density, and interaction tone before adapting copy or content.
Key visible headings include: Migrate to the cloud with confidence; Discover untethered scalability. The platform built to dynamically map, migrate, and optimize your entire infrastructure without lifting a finger.; Architectural Clarity; Dynamic Scaling.
## Colors
Anchor the palette in primary #2563EB, secondary #050510, accent #3B82F6, background #050510, surface #18181B, text-primary #FFFFFF. Keep background, surface, text, and border roles distinct so generated layouts retain the same contrast pattern as the source.
## Typography
Use Inter for display moments and Inter for body copy unless the HTML clearly demands a compatible fallback. Labels and technical metadata should use JetBrains Mono or an equivalent mono face.
## Layout
Keep spacing deliberate and stable. Favor the same grid direction, max-width behavior, card density, and responsive stacking seen in the HTML. Do not replace distinctive source structures with generic SaaS sections.
## Components
Dashboard, chart, and data panels should preserve their compact operational hierarchy, nested surfaces, and metric emphasis.
## Motion
Preserve existing motion cues such as masked reveals, staggered entrance, hover lift, scroll-triggered transitions, and ambient movement. Keep easing smooth and restrained.
## WebGL & Effects
If the source includes canvas, WebGL, Three.js, gradients, particles, or atmospheric effects, rebuild them as supporting layers behind the content. Keep effects performant, responsive, and secondary to the interface.
## Guardrails
- Do not flatten the source into a generic card grid.
- Do not swap the color mode unless the source clearly supports it.
- Preserve the first viewport signal, focal object, and visual density.
- Keep buttons, cards, and badges aligned to the same radius and border language.
| F1 | Exact-hash cache key menyertakan context (channel/thread) → teks sama di channel lain selalu miss | Killer hit-rate #1 |
| F2 | Semantic tier TIDAK memfilter context (Qdrant payload tak punya context) — sudah global tapi hanya aman krn sim 0.97 ketat | Inkonsisten dgn exact tier |
| F3 | Phase-1 lookup loop `await getCachedTextModeration(key)` per pesan → N round-trip PgBouncer per batch (60 msg = 60 query serial) | Latensi + beban DB |
| F4 | Verdict actionable (flagged/warn) dan clean sama-sama boleh di-serve semantic; toleransi akurasi beda | Risiko akurasi |
| F5 | `hit_count` tidak pernah di-increment oleh reader manapun | Hit-rate tak terukur |
| F6 | `wikipediaSearch()` (blok `<web_searches>`) tanpa cache — re-fetch tiap batch utk query sama | Latensi + spam ke WP |
| F7 | `fetchUrlSafely()` tanpa cache — link sama di batch berikutnya di-download lagi penuh | Latensi + bandwidth |
| F8 | Vision cache key dari data-URL base64 hasil resize → attachment sama via jalur berbeda (URL vs embed) = key beda → re-download + re-vision | Duplikasi kerja vision |
Non-goals: mengubah pipeline enforcement (auto-mute/ban trust-store writes), mengubah prompt
kebijakan moderasi, mengubah model/embedding provider.
## Desain
Semua perubahan degrade gracefully — cache gagal → perilaku lama (LLM). Akurasi dilindungi
asimetris: **hemat boleh untuk verdict non-actionable, konservatif untuk yang memicu aksi.**
### D1 — Cache metrics (F5)
-`textCacheStore.getCachedTextModeration()`: saat hit valid, increment `hit_count`
"Ambil skor trust, jumlah infraction, dan streak pesan bersih seorang user dari user_reputations. Untuk 'berapa trust score si A?' / riwayat pelanggaran. Butuh user_id.",
"Cek status skor reputasi per-user. CATATAN: fitur trust score per-user telah dihapus dari sistem — tool ini menjawab 'unavailable' dan menyarankan rasio flag vs total sebagai pengganti.",
parameters:{
type:"object",
properties:{
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
asepharyana
asepharyana
asepharyana
dependabot[bot]
mytheclipsebotreview[bot]
mytheclipsebotreview
Asep Hariyana
mytheclipsebotreview
