feat(api): OpenAPI spec update via Stainless API (#71)

src/resources/agent.ts

@@ -108,6 +108,24 @@ export interface AgentResponse {
     | 'static-noise'
     | null;
 
+  /**
+   * Only applicable when enable_backchannel is true. Controls how often the agent
+   * would backchannel when a backchannel is possible. Value ranging from [0,1].
+   * Lower value means less frequent backchannel, while higher value means more
+   * frequent backchannel. If unset, default value 0.8 will apply.
+   */
+  backchannel_frequency?: number;
+
+  /**
+   * Only applicable when enable_backchannel is true. A list of words that the agent
+   * would use as backchannel. If not set, default backchannel words will apply.
+   * Check out
+   * [backchannel default words](/agent/interaction-configuration#backchannel) for
+   * more details. Note that certain voices do not work too well with certain words,
+   * so it's recommended to expeirment before adding any words.
+   */
+  backchannel_words?: Array<string>;
+
   /**
    * Provide a customized list of keywords to bias the transcriber model, so that
    * these words are more likely to get transcribed. Commonly used for names, brands,
@@ -133,25 +151,9 @@ export interface AgentResponse {
   interruption_sensitivity?: number;
 
   /**
-   * `Beta feature, use with caution.`
-   *
-   * This setting specifies the agent's operational language, including base language
-   * and dialect. Speech recognition considers both elements, but text-to-speech
-   * currently only recognizes the base language.
-   *
+   * Specifies what language (and dialect) the speech recognition will operate in.
    * For instance, selecting `en-GB` optimizes speech recognition for British
-   * English, yet text-to-speech output will be in standard English. If
-   * dialect-specific text-to-speech is required, please contact us for support.
-   *
-   * If unset, will use default value `en-US`.
-   *
-   * - `11lab voices`: supports English(en), German(de), Spanish(es), Hindi(hi),
-   *   Portuguese(pt)
-   *
-   * - `openAI voices`: supports English(en), German(de), Spanish(es), Hindi(hi),
-   *   Portuguese(pt), Japanese(ja)
-   *
-   * - `deepgram voices`: supports English(en)
+   * English. If unset, will use default value `en-US`.
    */
   language?:
     | 'en-US'
@@ -173,6 +175,20 @@ export interface AgentResponse {
    */
   opt_out_sensitive_data_storage?: boolean;
 
+  /**
+   * If set, controls how many times agent would remind user when user is
+   * unresponsive. Must be a non negative integer. If unset, default value of 1 will
+   * apply (remind once). Set to 0 to disable agent from reminding.
+   */
+  reminder_max_count?: number;
+
+  /**
+   * If set (in milliseconds), will trigger a reminder to the agent to speak if the
+   * user has been silent for the specified duration after some agent speech. Must be
+   * a positive number. If unset, default value of 10000 ms (10 s) will apply.
+   */
+  reminder_trigger_ms?: number;
+
   /**
    * Controls how responsive is the agent. Value ranging from [0,1]. Lower value
    * means less responsive agent (wait more, respond slower), while higher value
@@ -256,6 +272,24 @@ export interface AgentCreateParams {
     | 'static-noise'
     | null;
 
+  /**
+   * Only applicable when enable_backchannel is true. Controls how often the agent
+   * would backchannel when a backchannel is possible. Value ranging from [0,1].
+   * Lower value means less frequent backchannel, while higher value means more
+   * frequent backchannel. If unset, default value 0.8 will apply.
+   */
+  backchannel_frequency?: number;
+
+  /**
+   * Only applicable when enable_backchannel is true. A list of words that the agent
+   * would use as backchannel. If not set, default backchannel words will apply.
+   * Check out
+   * [backchannel default words](/agent/interaction-configuration#backchannel) for
+   * more details. Note that certain voices do not work too well with certain words,
+   * so it's recommended to expeirment before adding any words.
+   */
+  backchannel_words?: Array<string>;
+
   /**
    * Provide a customized list of keywords to bias the transcriber model, so that
    * these words are more likely to get transcribed. Commonly used for names, brands,
@@ -281,25 +315,9 @@ export interface AgentCreateParams {
   interruption_sensitivity?: number;
 
   /**
-   * `Beta feature, use with caution.`
-   *
-   * This setting specifies the agent's operational language, including base language
-   * and dialect. Speech recognition considers both elements, but text-to-speech
-   * currently only recognizes the base language.
-   *
+   * Specifies what language (and dialect) the speech recognition will operate in.
    * For instance, selecting `en-GB` optimizes speech recognition for British
-   * English, yet text-to-speech output will be in standard English. If
-   * dialect-specific text-to-speech is required, please contact us for support.
-   *
-   * If unset, will use default value `en-US`.
-   *
-   * - `11lab voices`: supports English(en), German(de), Spanish(es), Hindi(hi),
-   *   Portuguese(pt)
-   *
-   * - `openAI voices`: supports English(en), German(de), Spanish(es), Hindi(hi),
-   *   Portuguese(pt), Japanese(ja)
-   *
-   * - `deepgram voices`: supports English(en)
+   * English. If unset, will use default value `en-US`.
    */
   language?:
     | 'en-US'
@@ -321,6 +339,20 @@ export interface AgentCreateParams {
    */
   opt_out_sensitive_data_storage?: boolean;
 
+  /**
+   * If set, controls how many times agent would remind user when user is
+   * unresponsive. Must be a non negative integer. If unset, default value of 1 will
+   * apply (remind once). Set to 0 to disable agent from reminding.
+   */
+  reminder_max_count?: number;
+
+  /**
+   * If set (in milliseconds), will trigger a reminder to the agent to speak if the
+   * user has been silent for the specified duration after some agent speech. Must be
+   * a positive number. If unset, default value of 10000 ms (10 s) will apply.
+   */
+  reminder_trigger_ms?: number;
+
   /**
    * Controls how responsive is the agent. Value ranging from [0,1]. Lower value
    * means less responsive agent (wait more, respond slower), while higher value
@@ -389,6 +421,24 @@ export interface AgentUpdateParams {
     | 'static-noise'
     | null;
 
+  /**
+   * Only applicable when enable_backchannel is true. Controls how often the agent
+   * would backchannel when a backchannel is possible. Value ranging from [0,1].
+   * Lower value means less frequent backchannel, while higher value means more
+   * frequent backchannel. If unset, default value 0.8 will apply.
+   */
+  backchannel_frequency?: number;
+
+  /**
+   * Only applicable when enable_backchannel is true. A list of words that the agent
+   * would use as backchannel. If not set, default backchannel words will apply.
+   * Check out
+   * [backchannel default words](/agent/interaction-configuration#backchannel) for
+   * more details. Note that certain voices do not work too well with certain words,
+   * so it's recommended to expeirment before adding any words.
+   */
+  backchannel_words?: Array<string>;
+
   /**
    * Provide a customized list of keywords to bias the transcriber model, so that
    * these words are more likely to get transcribed. Commonly used for names, brands,
@@ -414,25 +464,9 @@ export interface AgentUpdateParams {
   interruption_sensitivity?: number;
 
   /**
-   * `Beta feature, use with caution.`
-   *
-   * This setting specifies the agent's operational language, including base language
-   * and dialect. Speech recognition considers both elements, but text-to-speech
-   * currently only recognizes the base language.
-   *
+   * Specifies what language (and dialect) the speech recognition will operate in.
    * For instance, selecting `en-GB` optimizes speech recognition for British
-   * English, yet text-to-speech output will be in standard English. If
-   * dialect-specific text-to-speech is required, please contact us for support.
-   *
-   * If unset, will use default value `en-US`.
-   *
-   * - `11lab voices`: supports English(en), German(de), Spanish(es), Hindi(hi),
-   *   Portuguese(pt)
-   *
-   * - `openAI voices`: supports English(en), German(de), Spanish(es), Hindi(hi),
-   *   Portuguese(pt), Japanese(ja)
-   *
-   * - `deepgram voices`: supports English(en)
+   * English. If unset, will use default value `en-US`.
    */
   language?:
     | 'en-US'
@@ -461,6 +495,20 @@ export interface AgentUpdateParams {
    */
   opt_out_sensitive_data_storage?: boolean;
 
+  /**
+   * If set, controls how many times agent would remind user when user is
+   * unresponsive. Must be a non negative integer. If unset, default value of 1 will
+   * apply (remind once). Set to 0 to disable agent from reminding.
+   */
+  reminder_max_count?: number;
+
+  /**
+   * If set (in milliseconds), will trigger a reminder to the agent to speak if the
+   * user has been silent for the specified duration after some agent speech. Must be
+   * a positive number. If unset, default value of 10000 ms (10 s) will apply.
+   */
+  reminder_trigger_ms?: number;
+
   /**
    * Controls how responsive is the agent. Value ranging from [0,1]. Lower value
    * means less responsive agent (wait more, respond slower), while higher value

src/resources/call.ts

@@ -51,7 +51,7 @@ export interface CallResponse extends RegisterCallResponse {
    *   completion status and other metrics. Available after call ends. Subscribe to
    *   `call_analyzed` webhook event type to receive it once ready.
    */
-  conversation_eval?: CallResponse.ConversationEval;
+  call_analysis?: CallResponse.CallAnalysis;
 
   /**
    * The reason for the disconnection of the call. Debug using explanation in docs
@@ -150,39 +150,39 @@ export namespace CallResponse {
    *   completion status and other metrics. Available after call ends. Subscribe to
    *   `call_analyzed` webhook event type to receive it once ready.
    */
-  export interface ConversationEval {
+  export interface CallAnalysis {
     /**
-     * Evaluate agent task completion status, whether the agent has completed his task.
+     * Sentiment of the agent in the call.
      */
-    agent_task_completion?: 'Completed' | 'Incomplete' | 'Partial';
+    agent_sentiment?: 'Negative' | 'Positive' | 'Neutral';
 
     /**
-     * Reason for the agent task completion status.
+     * Evaluate agent task completion status, whether the agent has completed his task.
      */
-    agent_task_completion_reason?: string;
+    agent_task_completion_rating?: 'Complete' | 'Incomplete' | 'Partial';
 
     /**
-     * Sentiment of the agent in the conversation.
+     * Reason for the agent task completion status.
      */
-    agnet_sentiment?: 'Negative' | 'Positive' | 'Neutral';
+    agent_task_completion_rating_reason?: string;
 
     /**
-     * Evaluate whether the conversation ended normally or was cut off.
+     * Evaluate whether the call ended normally or was cut off.
      */
-    conversation_completion?: 'Completed' | 'Incomplete' | 'Partial';
+    call_completion_rating?: 'Complete' | 'Incomplete' | 'Partial';
 
     /**
-     * Reason for the conversation completion status.
+     * Reason for the call completion status.
      */
-    conversation_completion_reason?: string;
+    call_completion_rating_reason?: string;
 
     /**
-     * A high level summary of the conversation conversation.
+     * A high level summary of the call.
      */
-    conversation_summary?: string;
+    call_summary?: string;
 
     /**
-     * Sentiment of the user in the conversation.
+     * Sentiment of the user in the call.
      */
     user_sentiment?: 'Negative' | 'Positive' | 'Neutral';
   }
@@ -326,9 +326,9 @@ export namespace CallResponse {
     role: 'agent' | 'user';
 
     /**
-     * Array of words in the utternace with the word timestamp. Useful for
+     * Array of words in the utterance with the word timestamp. Useful for
      * understanding what word was spoken at what time. Note that the word timestamp is
-     * not guranteed to be accurate, it's more like an approximation.
+     * not guaranteed to be accurate, it's more like an approximation.
      */
     words: Array<TranscriptObject.Word>;
   }
@@ -366,9 +366,9 @@ export namespace CallResponse {
     role: 'agent' | 'user';
 
     /**
-     * Array of words in the utternace with the word timestamp. Useful for
+     * Array of words in the utterance with the word timestamp. Useful for
      * understanding what word was spoken at what time. Note that the word timestamp is
-     * not guranteed to be accurate, it's more like an approximation.
+     * not guaranteed to be accurate, it's more like an approximation.
      */
     words: Array<Utterance.Word>;
   }
@@ -689,10 +689,10 @@ export interface CallRegisterParams {
   from_number?: string;
 
   /**
-   * An abtriary object for storage purpose only. You can put anything here like your
-   * own id for the call, twilio SID, internal customer id. Not used for processing,
-   * when we connect to your LLM websocket server, you can then get it from the call
-   * object.
+   * An arbitrary object for storage purpose only. You can put anything here like
+   * your own id for the call, twilio SID, internal customer id. Not used for
+   * processing, when we connect to your LLM websocket server, you can then get it
+   * from the call object.
    */
   metadata?: unknown;
 

tests/api-resources/agent.test.ts

@@ -29,11 +29,15 @@ describe('resource agent', () => {
       voice_id: '11labs-Adrian',
       agent_name: 'Jarvis',
       ambient_sound: 'coffee-shop',
+      backchannel_frequency: 0.9,
+      backchannel_words: ['yeah', 'uh-huh'],
       boosted_keywords: ['retell', 'kroger'],
       enable_backchannel: true,
       interruption_sensitivity: 1,
       language: 'en-US',
       opt_out_sensitive_data_storage: true,
+      reminder_max_count: 2,
+      reminder_trigger_ms: 10000,
       responsiveness: 1,
       voice_speed: 1,
       voice_temperature: 1,