- docs/open-source/conversation-mechanics.mdx: StreamingConversation.__init__ parameter is speed_coefficient, not conversation_speed
- docs/open-source/create-your-own-agent.mdx: The `respond` method signature is missing `conversation_id: str` parameter, is not marked as `async`, and return statement is missing the boolean value. The `generate_response` method signature is missing `conversation_id: str` and `bot_was_in_medias_res: bool = False` parameters, and should yield `GeneratedResponse` instead of simplified `Tuple[str, bool]`. All `RespondAgent.respond` implementations include `conversation_id`.
- docs/open-source/language-support.mdx: Incorrect syntax in os.getenv (default= keyword arg invalid, must be positional); incorrect parameter name 'voice' instead of 'voice_name' for AzureSynthesizerConfig; incorrect use of 'model' parameter with language default value instead of 'language' parameter for DeepgramTranscriberConfig
- docs/open-source/python-quickstart.mdx: `create_streaming_microphone_input_and_speaker_output` does not accept `use_blocking_speaker_output` parameter; it always uses `BlockingStreamingSpeakerOutput`.
- docs/open-source/external-action.mdx: Response schema is missing the required `success: bool` field present in `ExternalActionResponse`. | `processing_mode` has no default value and is listed as required in class inputs, but doc implies optional by not listing default.
- docs/open-source/synthesizer-reference.mdx: Parameter type for speed is Optional[float] not Optional[int] | Parameter type for temperature is Optional[float] not Optional[int] | Parameter experimental_streaming is bool (required) not Optional[bool]
- docs/open-source/synthesizer.mdx: voice_name is str (required) not Optional[str] with required=false
- docs/open-source/transcriber-reference.mdx: DeepgramTranscriberConfig.model is Optional[str] with no default value specified | DeepgramTranscriberConfig.version is Optional[str] with no default value specified | GoogleTranscriberConfig.language_code is str (required) with no default value | AssemblyAITranscriberConfig.buffer_size_seconds is float (required) with no default value | WhisperCPPTranscriberConfig.buffer_size_seconds is float (required) with no default value | AzureTranscriberConfig.language is str (required) with no default value | GladiaTranscriberConfig.buffer_size_seconds is float (required) with no default value
- docs/open-source/langchain-agent-dep.mdx: TelephonyServer.__init__ does not accept a 'logger' parameter per the function signature | OutboundCall.__init__ requires 'telephony_config' parameter (no default) but it is missing; does not accept 'logger' parameter per the function signature
- docs/open-source/using-synthesizers.mdx: Example code uses from_telephone_output_device but description incorrectly states from_output_device
- docs/open-source/using-transcribers.mdx: Code imports PunctuationEndpointingConfig but uses DeepgramEndpointingConfig | Code imports PunctuationEndpointingConfig but uses DeepgramEndpointingConfig; from_input_device is on base TranscriberConfig