@pyai/sdk 0.17.0 → 0.18.0

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/src/index.ts CHANGED
@@ -129,7 +129,8 @@ export interface SpeechParams {
129
129
  voice?: string;
130
130
  model?: SpeakModel;
131
131
  /**
132
- * Delivery mode. The API defaults to true and every catalog voice accepts
132
+ * Delivery mode. `speech` defaults to false; `speechStream` always uses true.
133
+ * The API itself defaults to true and every catalog voice accepts
133
134
  * both values. Set false when you want one complete buffered body with a
134
135
  * `Content-Length`. A voice whose serving fleet has no streaming lane is
135
136
  * rendered buffered either way and says so with `x-pyai-stream: buffered`.
@@ -183,6 +184,10 @@ export type CreateJobParams = CreateJobOptions & (
183
184
  );
184
185
 
185
186
  export interface CreateJobOptions {
187
+ /** Spoken-language hint for every audio chunk; omit for automatic detection. */
188
+ transcription_language?: string;
189
+ /** Original script (default), or Roman letters preserving the spoken language; no translation. */
190
+ output_script?: "original" | "latin";
186
191
  /** Requires Trace entitlement; incompatible with channel/diarize. */
187
192
  trace?: boolean;
188
193
  rule_pack?: Record<string, unknown>;
@@ -500,7 +505,7 @@ export interface HearStreamOptions {
500
505
  /** Streaming STT model. Server default `pyai-hear`. */
501
506
  model?: string;
502
507
  /** Omit or use auto for automatic detection; an explicit code pins recognition. */
503
- language?: "auto" | "en" | "es" | "fr" | "de" | "hi" | "it" | "pt" | "nl" | "zh" | "yue" | "ar" | "id" | "ko" | "th" | "vi" | "ja" | "tr" | "ms" | "sv" | "da" | "fi" | "pl" | "cs" | "fil" | "fa" | "el" | "hu" | "mk" | "ro";
508
+ language?: "auto" | "en" | "es" | "fr" | "de" | "hi" | "it" | "pt" | "nl" | "zh" | "yue" | "ar" | "id" | "ko" | "th" | "vi" | "ja" | "tr" | "ms" | "sv" | "da" | "fi" | "pl" | "cs" | "fil" | "fa" | "el" | "hu" | "mk" | "ro" | "ru";
504
509
  /** Input PCM sample rate in Hz. Default 16000 server-side. */
505
510
  sampleRate?: number;
506
511
  /** Audio frame encoding. Default "pcm16". */
@@ -1650,6 +1655,21 @@ export interface TelephonyNumberPurchase {
1650
1655
  max_monthly_cost_cents?: number;
1651
1656
  }
1652
1657
 
1658
+ export interface TelephonyDesktopEndpoint {
1659
+ object:'telephony.desktop_endpoint';id:string;project_id:string;user_id:string;actor_type:'application';label:string;
1660
+ registered:boolean;available:boolean;number_id:string;status:'active'|'revoked';
1661
+ }
1662
+ export interface TelephonyDesktopEnrollment {object:'telephony.desktop_enrollment';version:1;endpoint_id:string;enrollment_token:string;expires_at:number;redeem_path:string;}
1663
+ export interface TelephonyDesktopCredentials {endpoint_id:string;org_id:string;project_id:string;user_id:string;label:string;refresh_token:string;refresh_expires_at:number;access_token:string;expires_at:number;presence_token:string;heartbeat_interval_seconds:15;}
1664
+ export interface TelephonyInboundRingGroup {
1665
+ mode:'simultaneous'|'desktop_first';desktop_endpoint_ids:[string];external_to:string;desktop_timeout_seconds:number;
1666
+ ring_timeout_seconds:number;on_no_answer:'hangup';max_duration_seconds:number;recording:boolean;
1667
+ }
1668
+ export interface TelephonyInboundCapabilities {
1669
+ object:'telephony.inbound_capabilities';number_id:string;binding_version:number;
1670
+ ring_group:{supported:boolean;modes:Array<'simultaneous'|'desktop_first'>;max_desktop_endpoints:1;on_no_answer:['hangup'];desktop_registration_required:true};
1671
+ }
1672
+
1653
1673
  export interface TelephonyInboundBinding {
1654
1674
  object: "telephony.inbound_binding";
1655
1675
  number_id: string;
@@ -1661,6 +1681,7 @@ export interface TelephonyInboundBinding {
1661
1681
  agent_profile_sha256?: string;
1662
1682
  admission_mode?: "client" | "platform";
1663
1683
  forward?: TelephonyInboundForward | null;
1684
+ ring_group?:TelephonyInboundRingGroup|null;
1664
1685
  carrier_operation?: { id: string; phase: "pending" | "unknown" | "active" | "rollback_pending" | "rolled_back"; action: "activate" | "rollback" };
1665
1686
  }
1666
1687
 
@@ -1693,6 +1714,7 @@ export type TelephonyInboundBindingUpdate = {
1693
1714
  admission_mode?: "client" | "platform";
1694
1715
  agent_knowledge_receipt?: Record<string, unknown>;
1695
1716
  forward?: TelephonyInboundForward | null;
1717
+ ring_group?:TelephonyInboundRingGroup|null;
1696
1718
  });
1697
1719
 
1698
1720
  /* ------------------------------------------------------------------------- *
@@ -2320,13 +2342,13 @@ export class PyAI {
2320
2342
  // --- audio --------------------------------------------------------------
2321
2343
 
2322
2344
  audio = {
2323
- /** Text-to-speech. Returns the raw audio bytes (default WAV). */
2345
+ /** Text-to-speech. Returns complete audio bytes (default WAV, stream:false). */
2324
2346
  speech: async (params: SpeechParams): Promise<ArrayBuffer> => {
2325
2347
  assertActiveSpeechParams(params);
2326
2348
  const res = await this.request("/v1/audio/speech", {
2327
2349
  method: "POST",
2328
2350
  headers: this.authHeaders({ "Content-Type": "application/json" }),
2329
- body: JSON.stringify({ model: "pyai-speak", ...params }),
2351
+ body: JSON.stringify({ model: "pyai-speak", ...params, stream: params.stream ?? false }),
2330
2352
  });
2331
2353
  await validateSpeechResponse(res, params);
2332
2354
  return res.arrayBuffer();
@@ -2417,11 +2439,15 @@ export class PyAI {
2417
2439
  },
2418
2440
  /** Upload a recording with a durable retry key. Requires the current
2419
2441
  * keyed multipart job contract; never change the key after an unknown ACK. */
2420
- upload: async (params: { file: Blob; filename: string; call_id?: string; diarize?: boolean; output_formats?: Array<"json" | "srt" | "vtt"> }, opts: { idempotencyKey: string }): Promise<TranscriptionJob> => {
2442
+ upload: async (params: { file: Blob; filename: string; call_id?: string; diarize?: boolean; channel?: boolean; transcription_language?: string; output_script?: "original" | "latin"; webhook_url?: string; output_formats?: Array<"json" | "srt" | "vtt"> }, opts: { idempotencyKey: string }): Promise<TranscriptionJob> => {
2421
2443
  if (!opts.idempotencyKey || opts.idempotencyKey.length > 255 || /[\r\n]/.test(opts.idempotencyKey)) throw new Error("A valid recording retry key is required.");
2422
2444
  const form = new FormData();
2423
2445
  if (params.call_id) form.set("call_id", params.call_id);
2424
2446
  if (params.diarize !== undefined) form.set("diarize", String(params.diarize));
2447
+ if (params.channel !== undefined) form.set("channel", String(params.channel));
2448
+ if (params.transcription_language) form.set("transcription_language", params.transcription_language);
2449
+ if (params.output_script) form.set("output_script", params.output_script);
2450
+ if (params.webhook_url) form.set("webhook_url", params.webhook_url);
2425
2451
  if (params.output_formats) form.set("output_formats", params.output_formats.join(","));
2426
2452
  form.set("audio", params.file, params.filename);
2427
2453
  const response = await this.request("/v1/transcription/jobs", { method: "POST", headers: this.authHeaders({ "Idempotency-Key": opts.idempotencyKey }), body: form });
@@ -2482,6 +2508,11 @@ export class PyAI {
2482
2508
 
2483
2509
  telephony = {
2484
2510
  /** Server-side methods. Keep the PyAI key in your backend. */
2511
+ desktopEndpoints: {
2512
+ list:(input:{number_id?:string}={}):Promise<ListResponse<TelephonyDesktopEndpoint>>=>this.getJson('/v1/telephony/desktop-endpoints'+(input.number_id?'?'+new URLSearchParams({number_id:input.number_id}):'')),
2513
+ create:(input:{number_id:string;user_id:string;client_reference:string;label:string}):Promise<TelephonyDesktopEndpoint>=>this.postJson('/v1/telephony/desktop-endpoints',input),
2514
+ enroll:(id:string,input:{user_id:string},options:{idempotencyKey:string}):Promise<TelephonyDesktopEnrollment>=>this.postJson(`/v1/telephony/desktop-endpoints/${encodeURIComponent(id)}/enrollment`,input,telephonyIdempotencyHeader(options.idempotencyKey)),
2515
+ },
2485
2516
  calls: {
2486
2517
  create: (input: TelephonyCallCreate, options: { idempotencyKey: string }): Promise<TelephonyCall> =>
2487
2518
  this.postJson("/v1/telephony/calls", input, telephonyIdempotencyHeader(options.idempotencyKey)),
@@ -2548,6 +2579,7 @@ export class PyAI {
2548
2579
  /** Route a number to an agent (`agentId: null` to unassign). */
2549
2580
  assign: (id: string, agentId: string | null): Promise<TelephonyNumber> =>
2550
2581
  this.postJson(`/v1/telephony/numbers/${encodeURIComponent(id)}/assign`, { agent_id: agentId }),
2582
+ getInboundCapabilities:(id:string):Promise<TelephonyInboundCapabilities>=>this.getJson(`/v1/telephony/numbers/${encodeURIComponent(id)}/inbound-capabilities`),
2551
2583
  getInboundBinding: (id: string): Promise<TelephonyInboundBinding> =>
2552
2584
  this.getJson(`/v1/telephony/numbers/${encodeURIComponent(id)}/inbound-binding`),
2553
2585
  /** Save the profile and stable operation first; after uncertainty read the