Use when evaluating, extending, or writing C# code for conversational speech-to-text, Flux-style real-time transcription, or turn-taking streaming in the Deepgram .NET SDK. Identifies missing Flux request parameters (language_hint, eot_threshold), maps existing WebSocket response types, provides the closest supported LiveSchema code path, and guides adding TurnInfo models and Flux examples. Use `deepgram-dotnet-speech-to-text` for standard streaming transcription without turn awareness.
94
93%
Does it follow best practices?
Impact
94%
1.25xAverage score across 3 eval scenarios
Passed
No known issues
Live streaming transcription with utterance detection
ClientFactory usage
100%
100%
Library.Initialize call
100%
100%
Correct LiveSchema namespace
100%
100%
Subscribe pattern
100%
100%
ResultResponse subscription
100%
100%
Transcript null/whitespace filter
71%
100%
UtteranceEndResponse subscription
100%
100%
UtteranceEnd parameter
75%
100%
VadEvents parameter
0%
100%
InterimResults parameter
100%
100%
No invented TurnInfo models
100%
100%
Flux SDK extension workflow
LiveSchema LanguageHint field
100%
100%
LiveSchema EotThreshold field
100%
100%
LiveSchema EagerEotThreshold field
100%
100%
TurnInfo model path
30%
100%
Client.cs dispatch updated
60%
100%
Unit tests for request serialization
88%
100%
Unit tests for response deserialization
88%
100%
Example at correct path
25%
0%
AsyncAPI spec reference
75%
100%
No ConnectFluxAsync invention
100%
100%
Workflow order respected
33%
55%
API version audit and long-session hardening
API version default identified
100%
100%
Connection URI logged
100%
100%
URI inspection before assuming /v2
70%
80%
SendKeepAlive used
46%
100%
SendFinalize used
100%
100%
SendClose used
25%
100%
No /v2/listen assumption
100%
100%
OpenResponse or CloseResponse subscribed
0%
100%
3b953f0
Table of Contents
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.