@absolutejs/voice-deepgram 0.0.18 → 0.0.20-beta.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -33,8 +33,8 @@ The adapter normalizes Deepgram events into the core `STTAdapter` contract.
33
33
  - `speech_final=true` -> normalized `endOfTurn`
34
34
  - `UtteranceEnd` -> normalized `endOfTurn`
35
35
  - Flux `EndOfTurn` or `EagerEndOfTurn` -> normalized `endOfTurn`
36
- - SDK transport errors -> normalized `error`
37
- - SDK socket close -> normalized `close`
36
+ - transport errors -> normalized `error`
37
+ - socket close -> normalized `close`
38
38
 
39
39
  ## Options
40
40
 
@@ -58,9 +58,18 @@ Supported options include:
58
58
  - `eagerEotThreshold`
59
59
  - `eotTimeoutMs`
60
60
  - `keepAliveMs`
61
+ - `connectTimeoutMs`
61
62
  - `tag`
62
63
  - `extra`
63
64
 
65
+ ### Diagnostics
66
+
67
+ `error` events now include `code` when available from Deepgram and include request identifiers when present.
68
+ If the websocket fails to authenticate or cannot open, the adapter throws with:
69
+ - transport details (or close reason)
70
+ - the websocket URL (query params only)
71
+ - effective timeout used for open handshake
72
+
64
73
  Nova models typically use `endpointing`, `utteranceEndMs`, and `vadEvents`.
65
74
 
66
75
  Flux models typically use:
@@ -69,6 +78,12 @@ Flux models typically use:
69
78
  - `eagerEotThreshold`
70
79
  - `eotTimeoutMs`
71
80
 
81
+ If you omit those Flux options, the adapter now applies conversation-oriented defaults:
82
+
83
+ - `eagerEotThreshold: 0.8`
84
+ - `eotThreshold: 0.82`
85
+ - `eotTimeoutMs: 1200`
86
+
72
87
  ## API Key
73
88
 
74
89
  Set `DEEPGRAM_API_KEY` in your runtime environment, or pass the key explicitly in the adapter config.
@@ -79,4 +94,4 @@ Deepgram references used for this adapter design:
79
94
  - https://developers.deepgram.com/docs/utterance-end
80
95
  - https://developers.deepgram.com/docs/endpointing
81
96
  - https://developers.deepgram.com/docs/flux/configuration
82
- - https://www.npmjs.com/package/%40deepgram/sdk
97
+ - https://developers.deepgram.com/docs/listening-to-audio-streaming-over-websocket