patrick-motsch
269b704812
Teamsbot: lower audio RMS threshold, add 10s min cooldown for name-based AI triggers
...
Made-with: Cursor
2026-02-27 22:38:29 +01:00
patrick-motsch
0ee86070bd
Teamsbot: fix chatHistory polluting AI context, block bot self-trigger via captions
...
Made-with: Cursor
2026-02-27 22:14:57 +01:00
patrick-motsch
681c96be8e
Teamsbot: prevent double AI triggers, remove caption text from AI context
...
Made-with: Cursor
2026-02-27 16:40:08 +01:00
patrick-motsch
275f7bbc66
Teamsbot: handle chatHistory source - store context but skip AI trigger
...
Made-with: Cursor
2026-02-27 13:56:46 +01:00
patrick-motsch
21f77d1924
Teamsbot: emit caption/speakerHint to UI for speaker display
...
Made-with: Cursor
2026-02-27 12:27:02 +01:00
patrick-motsch
1c00ab6024
Forward bot playback acknowledgements to SSE status stream.
...
Handle ttsPlaybackAck websocket messages and emit ttsDeliveryStatus updates with playback_* states for UI diagnostics.
Made-with: Cursor
2026-02-26 21:45:47 +01:00
patrick-motsch
25854edb4d
Allow speaker-hint captions to trigger AI without transcript persistence.
...
Keep aggressive hybrid persistence rules, but use final speaker hints as transient context to react to direct bot addressing even when audio capture is silent.
Made-with: Cursor
2026-02-26 21:41:56 +01:00
patrick-motsch
fe1a97564b
Improve Teams bot response reliability and transcript quality.
...
Fix invalid bot-response timestamps in SSE payloads, reduce duplicate response loops, and improve audio STT stability with larger capture chunks and safer silence filtering.
Made-with: Cursor
2026-02-26 21:18:21 +01:00
patrick-motsch
cd2fffc651
feat: emit ttsDeliveryStatus SSE diagnostics
...
Made-with: Cursor
2026-02-26 09:46:05 +01:00
patrick-motsch
02002f3576
feat: add speaker-hint debug flag and improve TTS diagnostics
...
Made-with: Cursor
2026-02-26 09:26:59 +01:00
patrick-motsch
0cd8e9ebfa
feat: treat caption events as speaker hints only
...
Made-with: Cursor
2026-02-26 09:22:07 +01:00
patrick-motsch
43e310a41d
feat: map audio STT to recent caption speaker hints
...
Made-with: Cursor
2026-02-26 09:05:21 +01:00
patrick-motsch
ea8cffee3c
fix: log teamsbot audio capture diagnostics in gateway
...
Read and log per-chunk capture diagnostics (trackId, readyState, rms, nativeSampleRate) from bot audioChunk messages to support root-cause analysis of silent audio streams.
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-24 14:55:43 +01:00
patrick-motsch
d020293f4b
fix: teamsbot audio STT - pass channels and skipFallbacks for known PCM16 format
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-23 22:59:53 +01:00
patrick-motsch
f6f42d8db7
phase 2 codeeditor and hotfixes voice
2026-02-23 22:09:27 +01:00
patrick-motsch
9d129f111f
fixes for roles
2026-02-23 17:13:50 +01:00
patrick-motsch
9c23fc6c52
fix: CSRF exemption for bot endpoints, WS ping timeout, clean up debug logs
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-19 00:46:42 +01:00
patrick-motsch
2d7da8a66d
fix: resolve STT AttributeError and int/str TypeError in teamsbot service
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-19 00:09:00 +01:00
patrick-motsch
7778325e5e
feat: add debugMode to config, filter bot own captions to prevent repeats
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-18 23:52:51 +01:00
patrick-motsch
b616688411
feat: add screenshot proxy endpoints for debug viewer (SysAdmin only)
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-18 22:00:34 +01:00
patrick-motsch
3fa5b98f47
security: restrict system bot access to SysAdmin only
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-18 21:03:08 +01:00
patrick-motsch
1227324703
feat: differential transcript, fix bot name in WS handler, AI commands support
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-18 17:50:31 +01:00
patrick-motsch
6963719499
fix: use enum .value for responseChannel comparison - was silently dropping all responses
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-18 00:48:41 +01:00
patrick-motsch
a3b771b12a
fix: derive botName from system bot email, robust responseChannel comparison
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-18 00:32:16 +01:00
patrick-motsch
33daa8dc74
feat: handle voiceGreeting message type - TTS for bot join greeting
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-18 00:17:42 +01:00
patrick-motsch
5d40657aef
fix: dynamic bot name in SPEECH_TEAMS prompt + phonetic name trigger
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-18 00:12:48 +01:00
patrick-motsch
070e3cbd96
fix: safe access to botAccountEmail in _isBotSpeaker - was crashing WebSocket handler on first transcript
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-17 23:54:18 +01:00
patrick-motsch
a27307263f
fix: create SSE event queue on-demand so transcript/chat events are never silently dropped
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-17 23:44:12 +01:00
patrick-motsch
a3c92ae8d5
fix: load system bot credentials from DB and pass to browser bot for authenticated join
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-17 19:58:01 +01:00
patrick-motsch
ad254aafb1
feat: handle audioChunk messages from bot, STT via Google Cloud Speech
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-17 18:43:42 +01:00
patrick-motsch
de573fd834
refactor: add TransferMode, remove backgroundImageUrl and botAccount fields from config
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-17 18:34:24 +01:00
patrick-motsch
5dd3b9f894
fix: add per-variant test-auth endpoints to avoid Azure timeout
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-17 16:33:21 +01:00
patrick-motsch
cd9b405d2e
fix: increase test-auth timeout to 900s for 5 browser variants
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-17 13:46:23 +01:00
patrick-motsch
7b4be3dbc9
feat: accept botEmail/botPassword from request, auto-create system bot in DB
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-17 00:33:05 +01:00
patrick-motsch
f743932768
fix: fallback to any active system bot if mandate mismatch, add credentialDebug to response
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-17 00:03:01 +01:00
patrick-motsch
1b2ca9512f
debug: add logging to test-auth credential loading
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-16 23:45:47 +01:00
patrick-motsch
bddb039bde
docs(teamsbot): update test-auth docstring variant count
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-16 23:06:56 +01:00
patrick-motsch
b08355ee45
teams test auth bot
2026-02-16 21:37:37 +01:00
patrick-motsch
3f9d791688
disable auth: use system bot display name, do not send credentials to browser bot
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-16 20:37:09 +01:00
patrick-motsch
a67c7f8b81
feat(teamsbot): stop via AI intent (multilingual), initial session context summarization
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-16 10:03:21 +01:00
patrick-motsch
4186ec6188
feat(teamsbot): stop command detection, session context for AI, context summarization for long sessions
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-16 09:29:03 +01:00
patrick-motsch
856b9f3c05
feat(teamsbot): system bot accounts with encrypted credentials, per-user settings, credentials removed from UI
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-16 00:16:02 +01:00
patrick-motsch
ef813a9304
feat(teamsbot): chat message handling, response channel (voice/chat/both), join mode (systemBot/anonymous/userAccount)
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-16 00:07:42 +01:00
patrick-motsch
cc86b144ac
fix(teamsbot): filter bot's own captions from AI trigger, mark bot responses in context to prevent repetition
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-15 22:52:02 +01:00
patrick-motsch
02dedc972d
fix(teamsbot): SSE rate-limit 60/min, TTS message format for bot protocol
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-15 15:19:31 +01:00
patrick-motsch
729c6355b5
fix: increase bot HTTP timeout to 60s for auth flow
...
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-15 12:20:58 +01:00
patrick-motsch
ad5c9d10cd
feat(teamsbot): dedicated bot account support with authenticated join
...
- New config fields: botAccountEmail, botAccountPassword for dedicated MSFT account
- BrowserBotConnector passes credentials + backgroundImageUrl to bot service
- Service passes config credentials to connector in joinMeeting
- Enables: full language settings, virtual background, no lobby wait
- Fallback: anonymous join when no bot account configured
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-15 11:56:04 +01:00
patrick-motsch
91425809c3
fix(teamsbot): AI prompt priority rules, bot-name trigger overrides cooldown
...
- SPEECH_TEAMS prompt: Rule 1 (bot name mentioned) now has highest priority
and ALWAYS triggers a response, even for greetings/smalltalk
- Bot-name detection in transcript overrides cooldown timer (immediate trigger)
- Added multilingual instruction: AI responds in the language it's addressed in
- Voice interface: clean camelCase mapping (audio_content -> audioContent)
- Config save: added "config" to allowed update fields in interfaceFeatures
- Default TTS voices for common languages in connectorVoiceGoogle
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-15 11:48:44 +01:00
patrick-motsch
e01c6dcb95
feat(teamsbot): HTTP fallback for transcripts, debug logging for WebSocket
...
- HTTP POST endpoints: /bot/transcript/{sessionId} and /bot/status/{sessionId}
- Fallback for when Azure App Service blocks WebSocket upgrade
- Debug logging [WS-DEBUG] in WebSocket handler for message tracking
- Handle websocket=None in AI response pipeline (HTTP mode)
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-15 11:05:58 +01:00
patrick-motsch
e24ef42617
feat(teamsbot): AI voice test, config save, camelCase mapping, default voices
...
- Voice test endpoint generates sample text dynamically via AI in selected language
- Fixed config save: added "config" to allowed update fields in interfaceFeatures
- Clean camelCase mapping in interfaceVoiceObjects (audio_content -> audioContent)
- Default TTS voices for common languages in connectorVoiceGoogle
- Fixed updateFeatureInstanceConfig -> updateFeatureInstance with config field
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-15 10:08:50 +01:00