ben.carrasco/jambonz-feature-server
1
0
Fork 0
mirror of https://github.com/jambonz/jambonz-feature-server.git synced 2025-12-20 08:40:38 +00:00
Code Issues Packages Projects Releases Wiki Activity
1,052 Commits 152 Branches 330 Tags
76df58bfc2cfd8420ffae446a2fbcf730fa0e1a0
Commit Graph

344 Commits

Author SHA1 Message Date
Dave Horton
76df58bfc2 fix logging in start task msg (#1202) 
* fix logging in start task msg

* generate uuids using native crypto lib
 2025-05-16 16:54:25 -04:00
Dave Horton
ca0bf36815 dont apply snake casing to either env vars or tool call args (#1194) (#1197) 2025-05-12 12:56:58 -04:00
rammohan-y
5c8237b382 Feat 1179 race issue with play verb (#1183) 
* Fixed race issue between queueCommand false and queueCommand true when play task is involved

https://github.com/jambonz/jambonz-feature-server/issues/1179

* removed unnecessary emitter

* added destroy mechanism for stickyEventEmitter

* clearing stickyEventEmitter

* memory leak fix
 2025-05-11 20:25:48 -04:00
Hoan Luu Huu
8b18532f31 fixed tts streaming buffer cannot reset timeoutwhen lastUpdateTime is short (#1184) 
* fixed tts streaming buffer cannot reset timeoutwhen lastUpdateTime is short

* wip
 2025-05-07 10:26:11 -04:00
Hoan Luu Huu
14295dcebc support google s2s (#1169) 
* support google s2s

* wip
 2025-05-07 07:20:33 -04:00
Hoan Luu Huu
9d54ca8116 Jambonz support Model context protocol (MCP) (#1150) 
* Jambonz support Model context protocol (MCP)

* merged mcp tools with existing llmOptions.tools

* support list of mcp servers

* wip

* wip

* wip

* fix voice agent

* fix open-ai

* fix review comment

* fix deepgram voice agent

* update verb specification version
 2025-04-24 06:50:53 -04:00
Sam Machin
472f4f4532 clientTools over webhooks (#1167) 
* clientTools over webhooks

* lint

* simpler toolHook response
 2025-04-23 09:15:16 -04:00
Hoan Luu Huu
15b583ef2c only connect to drachtio server if connected to freeswitch (#1123) 
* only connect to drachtio server if connected to freeswitch

* wip

* wip
 2025-04-22 09:55:39 -04:00
rammohan-y
182c310191 remove video from sdp in case of reInvite if the call is audio call (#1159) 
https://github.com/jambonz/jambonz-feature-server/issues/1158
 2025-04-18 09:33:16 -04:00
Hoan Luu Huu
343b382373 fixed ws-requestor missing hook for dial:confirm (#1143) 2025-04-09 07:29:08 -04:00
rammohan-y
d910981b1a Allow hangup verb on siprec call (#1136) 
* Allow hangup verb on siprec call
https://github.com/jambonz/jambonz-feature-server/issues/1135

* added sip:decline to AllowedSipRecVerbs
 2025-04-04 08:23:39 -04:00
Dave Horton
fcaf2e59e7 initial changes for openai stt (#1127) 
* initial changes for openai stt

* wip

* wip

* wip

* wip

* wip

* make minBargeinWordCount work for openai

* wip

* wip

* wip

* wip

* wip

* wip

* wip

* wipp

* wip

* wip

* wip

* openai stt: support for prompt templates

* lint

* wip

* support openai semantic_vad

* wip

* transcribe supports openai stt

* sip

* wip

* wip

* refactor list of stt vendors that dont need to be restarted after a final transcript

* cleanup

* wip

* cleanup

* wip

* wip

* wip

* remove credentials from log

* comment
 2025-03-28 13:14:58 -04:00
rammohan-y
ee846b283d Feat 1120 video call - remove video media from SDP if the call is audio call (#1124) 
* sending jambonz:error when the incoming message is not parsable

https://github.com/jambonz/jambonz-feature-server/issues/1094

* writing an alert when incoming paylod is invalid

* added content to the jambonz:error payload

* removing video media from sdp if the call is an audio call. This is to avoid sending video media to destination if the incoming call is an audio call

* calling removeVideoSdp only when the environment variable JAMBONES_VIDEO_CALLS_ENABLED_IN_FS is set to true, this will ensure there are no regression issues for audio calls

* fixed jslint errors
 2025-03-28 12:56:08 -04:00
Sam Machin
9c437ab687 use deepgramOptions.model (#1126) 
* use deepgramOptions.model

* lint

* Update transcription-utils.js
 2025-03-24 12:25:29 -04:00
rammohan-y
d36e6b4c22 set the detected language as language_code when deepgram detects the language (#1116) 
https://github.com/jambonz/jambonz-feature-server/issues/1115
 2025-03-11 12:16:29 -04:00
Sam Machin
3120dbc3e0 Feature: add digitCount to amd-utils (#1111) 
* add digitCount to amd-utils

* linting

* bump verb-specs
 2025-03-06 12:01:51 -05:00
Hoan Luu Huu
8b8283e603 ws requestor should store initial sessionData when sending session:adulting (#1110) 2025-03-06 07:42:47 -05:00
Dave Horton
bf4ae5b618 #1101 - allow listen url to have relative url and use base url of app… (#1102) 
* #1101 - allow listen url to have relative url and use base url of application if ws

* remove logging
 2025-02-28 14:19:45 -05:00
Hoan Luu Huu
af4e17f447 fixed dial transcribe is not able to receive final transcribe when closing the call (#1073) 
* fixed dial transcribe is not able to received final transcribe when close call.

* wip

* fix review comment

* support call session delay detroy ep when current task is transcribe

* wip

* wip

* fixed review comments

* fixed review comments
 2025-02-27 07:25:01 -05:00
Sam Machin
af475cbea4 Update place-outdial.js (#1090) 
* Update place-outdial.js

* update baseURL on redirect

* Revert "update baseURL on redirect"

This reverts commit 55778ba37edf029fa8687cd971b202af15478f95.
 2025-02-25 15:09:21 -05:00
rammohan-y
8bed44cce3 sending jambonz:error when the incoming message is not parsable (#1095) 
* sending jambonz:error when the incoming message is not parsable

https://github.com/jambonz/jambonz-feature-server/issues/1094

* writing an alert when incoming paylod is invalid

* added content to the jambonz:error payload
 2025-02-24 12:44:25 -05:00
Dave Horton
8ede41714b fix typo: change AWS_SNS_TOPIC_ARM to AWS_SNS_TOPIC_ARN (#1093) 2025-02-24 10:51:07 -05:00
Dave Horton
122d267816 better handling of flush commands (#1081) 
* better handling of flush commands

* rework buffering of tokens

* gather: when returning low confidence also provide the transcript

* better error handling in tts:tokens

* special handling of asr timeout for speechmatics

* remove some logs that were excessively wordy
 2025-02-18 09:31:11 -05:00
Hoan Luu Huu
7fa0041f6b support deepgram options noDelay (#1083) 
* support deepgram options noDelay

* update verb specification version
 2025-02-15 16:39:30 -05:00
Dave Horton
55b408eecb add support for deepgram keyterms (#1071) 2025-02-07 13:12:25 -05:00
Hoan Luu Huu
dbdc1cd43d support voxist stt (#1066) 
* support voxist stt

* wip
 2025-02-05 08:33:35 -05:00
Hoan Luu Huu
2ddcd53d6b support elevenlabs s2s (#1052) 
* support elevenlabs s2s

* wip

* wip

* wip
 2025-02-02 10:29:48 -05:00
rammohan-y
a4d07ddce0 Feat/1057 recognition mode (#1060) 
* sending recognition mode channel variable

* change verb-specifications version
 2025-01-28 08:06:04 -05:00
Hoan Luu Huu
ba282d775d support rimelabs tts streaming (#1047) 2025-01-18 08:17:33 -05:00
Dave Horton
a194ba833e Feat/1041 (#1045) 
* initial changes for stream synonym to listen

* listen on B endpoint if nested listen in dial has channel === 2
 2025-01-17 08:48:39 -05:00
rammohan-y
77f3d9d7ec feat/1034: sending socket close code when there is no response from the websocket app (#1035) 2025-01-16 10:13:00 -05:00
Sam Machin
4dbc7df93d new error for HTTP responses without stack trace (#1044) 
* new error for HTTP responses without stack trace

* lint
 2025-01-16 08:05:17 -05:00
Dave Horton
f71f0ac69a Fix/speechmatics (#1042) 
* add speechmatics options

* wip

* speechmatics does not do endpointing for us so we need to flip on continuousAsr

* speechmatics: continousAsr should be at least equal to max_delay, if set
 2025-01-15 19:12:15 -05:00
Hoan Luu Huu
ca8cace284 support custom tts streaming (#1023) 
* support custom tts streaming

* wip

* wip

* wip

* wip

* wip

* wip

* fix review comments
 2025-01-14 07:24:06 -05:00
Hoan Luu Huu
499c800213 Feat/ultravox s2s (#1032) 
* support ultravox_s2s

* support ultravox_s2s

* support ultravox_s2s

* wip

* wip

* wip

* wip

* fix ultravox toolcall

* wip
 2025-01-14 07:11:55 -05:00
Sam Machin
97952afb1d add deepgram filler words (#1036) 
* add deepgram filler words

* Update package.json

* Update package-lock.json
 2025-01-13 11:07:24 -05:00
Hoan Luu Huu
fcefa1ff31 fix inband dtmf does not work in dial verb (#1018) 2025-01-08 18:29:43 -05:00
Dave Horton
a2581eaeb4 tts throttling and send user_interruption event (#1019) 
* tts throttling and send user_interruption event

* tts streaming: if we get a flush with tokens pending, send the flush after the tokens

* wip
 2025-01-04 16:34:01 -05:00
Dave Horton
3706aa4d98 #1020 - fix for sticky bargein (#1021) 2025-01-03 10:41:35 -05:00
Dave Horton
25f1e65f63 feed TTS in sentence chunks when streaming (#1013) 
* feed TTS in sentence chunks when streaming

* tts streaming: treat a paragraph as a chunk of text, even it not ending with a line end character

* wip
 2024-12-31 15:16:25 -05:00
Hoan Luu Huu
564f6c9e55 support kill dial if sd ep is media timeout (#1001) 
* support kill dial if sd ep is media timeout

* support kill dial if sd ep is media timeout

* support kill dial if sd ep is media timeout

* add media timeout reason header to bye message

* wip

* wip

* make configuration for freeswitch media timeout

* make configuration for freeswitch media timeout

* wip
 2024-12-23 07:19:41 -05:00
Dave Horton
02f25f8343 fix cartesia channel vars for streaming (#1012) 2024-12-20 16:48:20 -05:00
Hoan Luu Huu
17250f8386 support cartesia tts (#1008) 
* support cartesia tts

* update speech util version

* update speech utils version
 2024-12-19 07:35:47 -05:00
Dave Horton
ba3f46df64 Feat/tts streaming (#994) 
* wip

* add TtsStreamingBuffer class to abstract handling of streaming tokens

* wip

* add throttling support

* support background ttsStream (#995)

* wip

* add TtsStreamingBuffer class to abstract handling of streaming tokens

* wip

* support background ttsStream

* wip

---------

Co-authored-by: Dave Horton <daveh@beachdognet.com>

* wip

* dont send if we have nothing to send

* initial testing with cartesia

* wip

---------

Co-authored-by: Hoan Luu Huu <110280845+xquanluu@users.noreply.github.com>
 2024-12-18 14:44:37 -05:00
RJ Burnham
f37e1540ee Make voicemail hints case insensitive (#1007) 2024-12-13 13:42:29 -05:00
Dave Horton
5e04db82bf Feat/deepgram voice agent (#1006) 
* wip

* wip

* wip
 2024-12-13 10:05:23 -05:00
Hoan Luu Huu
a929a649f9 fix ConfirmCallSession cannot be played (#993) 
* fix ConfirmCallSession cannot be played

* fix review comments

* fix review comments
 2024-12-10 19:36:42 -05:00
Dave Horton
3bb4f1a29f fix #998 incorrectly sending final transcript with is_final=false (#999) 2024-12-10 18:48:02 -05:00
rammohan-y
0458bb7d6c Feat/884: Capture system_alert when feature-server is online or offline (#950) 
* writing alerts during startup and shutdown of feature-server

* feat/884: created constants for system component name and state

* feat/88: added 0.2.11 version of time-series

* feat/884: renamed constant, and added GracefulShutdownInProgress system alert
 2024-12-05 09:23:03 -05:00
Sam Machin
dce4fe1f82 Fix/986 (#990) 
* throw new NonFatalTask error on play file not found

* linting

* make SpeechCredentialError subclass of NonFatalTask error

* cleanup

* Update action-hook-delay.js

* bump fsmrf version

* linting and package-lock

* Update package-lock.json

* update error

* only throw on fs error "File not found"

* add alert

* update time-series dep

* Update package-lock.json

* linting

* Update play.js

* remove stack trace from error message

* fix error formatting
 2024-12-04 05:47:49 -05:00