@fonoster/autopilot 0.7.4 → 0.7.6

package/dist/voice/Voice.js

@@ -0,0 +1,81 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.VoiceImpl = void 0;
+/*
+ * Copyright (C) 2024 by Fonoster Inc (https://fonoster.com)
+ * http://github.com/fonoster/fonoster
+ *
+ * This file is part of Fonoster
+ *
+ * Licensed under the MIT License (the "License");
+ * you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ *    https://opensource.org/licenses/MIT
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+const common_1 = require("@fonoster/common");
+const uuid_1 = require("uuid");
+class VoiceImpl {
+    constructor(sessionRef, voice) {
+        this.voice = voice;
+        this.sessionRef = sessionRef;
+        this.playbackRef = (0, uuid_1.v4)();
+    }
+    async answer() {
+        await this.voice.answer();
+    }
+    async hangup() {
+        await this.voice.hangup();
+    }
+    async say(text) {
+        await this.voice.say(text, { playbackRef: this.playbackRef });
+    }
+    async sgather() {
+        const stream = await this.voice.sgather({
+            source: common_1.StreamGatherSource.SPEECH
+        });
+        return {
+            stop: async () => {
+                stream.close();
+                stream.cleanup(() => { });
+            },
+            onData: (cb) => {
+                stream.onPayload((payload) => {
+                    cb(payload.speech);
+                });
+            }
+        };
+    }
+    async stream() {
+        const stream = await this.voice.stream();
+        return {
+            stop: async () => {
+                stream.close();
+                stream.cleanup(() => { });
+            },
+            onData: (cb) => {
+                stream.onPayload((payload) => {
+                    cb(payload.data);
+                });
+            }
+        };
+    }
+    async transfer(to, options) {
+        const { record, timeout } = options;
+        const effectiveOptions = {
+            recordDirection: record ? common_1.DialRecordDirection.BOTH : undefined,
+            timeout
+        };
+        await this.voice.dial(to, effectiveOptions);
+    }
+    async stopSpeech() {
+        await this.voice.playbackControl(this.playbackRef, common_1.PlaybackControlAction.STOP);
+    }
+}
+exports.VoiceImpl = VoiceImpl;

package/dist/voice/index.d.ts

@@ -0,0 +1,2 @@
+export * from "./Voice";
+export * from "./types";

package/dist/voice/index.js

@@ -0,0 +1,36 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __exportStar = (this && this.__exportStar) || function(m, exports) {
+    for (var p in m) if (p !== "default" && !Object.prototype.hasOwnProperty.call(exports, p)) __createBinding(exports, m, p);
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+__exportStar(require("./Voice"), exports);
+/*
+ * Copyright (C) 2024 by Fonoster Inc (https://fonoster.com)
+ * http://github.com/fonoster/fonoster
+ *
+ * This file is part of Fonoster
+ *
+ * Licensed under the MIT License (the "License");
+ * you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ *    https://opensource.org/licenses/MIT
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+__exportStar(require("./types"), exports);

package/dist/voice/types.d.ts

@@ -0,0 +1,23 @@
+type GatherStream = {
+    stop: () => Promise<void>;
+    onData: (cb: (speech: string) => void) => void;
+};
+type Stream = {
+    stop: () => Promise<void>;
+    onData: (cb: (chunk: Uint8Array) => void) => void;
+};
+type TransferOptions = {
+    timeout?: number;
+    record?: boolean;
+};
+type Voice = {
+    sessionRef: string;
+    answer: () => Promise<void>;
+    hangup: () => Promise<void>;
+    say: (text: string) => Promise<void>;
+    sgather: () => Promise<GatherStream>;
+    transfer: (destination: string, options?: TransferOptions) => Promise<void>;
+    stream: () => Promise<Stream>;
+    stopSpeech: () => Promise<void>;
+};
+export { TransferOptions, Voice };

package/dist/voice/types.js

@@ -0,0 +1,2 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });

package/dist/voiceServerSetup.d.ts

@@ -0,0 +1,2 @@
+declare function startVoiceServer(skipIdentity: boolean): void;
+export { startVoiceServer };

package/dist/{runner.js → voiceServerSetup.js}

@@ -1,9 +1,9 @@
-#!/usr/bin/env node
 "use strict";
 var __importDefault = (this && this.__importDefault) || function (mod) {
     return (mod && mod.__esModule) ? mod : { "default": mod };
 };
 Object.defineProperty(exports, "__esModule", { value: true });
+exports.startVoiceServer = startVoiceServer;
 /*
  * Copyright (C) 2024 by Fonoster Inc (https://fonoster.com)
  * http://github.com/fonoster/fonoster
@@ -22,25 +22,8 @@ Object.defineProperty(exports, "__esModule", { value: true });
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
-const logger_1 = require("@fonoster/logger");
 const voice_1 = __importDefault(require("@fonoster/voice"));
-const assistants_1 = require("./assistants");
-const Autopilot_1 = require("./Autopilot");
-const envs_1 = require("./envs");
-const logger = (0, logger_1.getLogger)({ service: "autopilot", filePath: __filename });
-const skipIdentity = process.env.NODE_ENV === "dev";
-new voice_1.default({ skipIdentity }).listen(async (req, voice) => {
-    const { ingressNumber, sessionRef, appRef } = req;
-    logger.verbose("voice request", { ingressNumber, sessionRef, appRef });
-    const assistantPath = process.argv[2];
-    const assistant = (0, assistants_1.loadAndValidateAssistant)(assistantPath);
-    logger.verbose("interacting with assistant", { name: assistant.name });
-    const autopilot = new Autopilot_1.Autopilot({
-        voice,
-        assistantConfig: {
-            ...assistant,
-            apiKey: envs_1.OPENAI_API_KEY
-        }
-    });
-    autopilot.start();
-});
+const handleVoiceRequest_1 = require("./handleVoiceRequest");
+function startVoiceServer(skipIdentity) {
+    new voice_1.default({ skipIdentity }).listen(handleVoiceRequest_1.handleVoiceRequest);
+}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@fonoster/autopilot",
-  "version": "0.7.4",
+  "version": "0.7.6",
   "description": "Voice AI for the Fonoster platform",
   "author": "Pedro Sanders <psanders@fonoster.com>",
   "homepage": "https://github.com/fonoster/fonoster#readme",
@@ -14,7 +14,8 @@
   "scripts": {
     "prebuild": "rimraf ./dist tsconfig.tsbuildinfo",
     "build": "tsc -b tsconfig.json",
-    "clean": "rimraf ./dist node_modules tsconfig.tsbuildinfo"
+    "clean": "rimraf ./dist node_modules tsconfig.tsbuildinfo",
+    "generate:readme": "node ../../.scripts/gen-readme.js"
   },
   "bin": {
     "autopilot": "./dist/runner.js"
@@ -34,17 +35,24 @@
     "url": "https://github.com/fonoster/fonoster/issues"
   },
   "dependencies": {
-    "@fonoster/logger": "^0.7.4",
-    "@fonoster/voice": "^0.7.4",
-    "@langchain/openai": "^0.2.7",
+    "@fonoster/common": "^0.7.5",
+    "@fonoster/logger": "^0.7.5",
+    "@fonoster/voice": "^0.7.5",
+    "@langchain/community": "^0.2.31",
+    "@langchain/core": "^0.2.31",
+    "@langchain/groq": "^0.0.17",
+    "@langchain/openai": "^0.2.8",
+    "cheerio": "^1.0.0",
     "dotenv": "^16.4.5",
+    "langchain": "^0.2.17",
     "onnxruntime-node": "^1.19.0",
+    "pdf-parse": "^1.1.1",
+    "uuid": "^10.0.0",
     "xstate": "^5.17.3",
     "zod": "^3.23.8"
   },
   "devDependencies": {
-    "node-record-lpcm16": "^1.0.1",
     "typescript": "^5.5.4"
   },
-  "gitHead": "e550aa46c1a9087a70157496365b64afd5aea11d"
+  "gitHead": "a02de2f037aaa9bc197e97c147ac9affe6cc69bb"
 }

package/dist/assistants/assistants.d.ts

@@ -1,4 +0,0 @@
-import { AssistantConfig } from "./types";
-type Assistant = ReturnType<typeof makeAssistant>;
-declare function makeAssistant(config: AssistantConfig): import("@langchain/core/runnables").Runnable<any, string, import("@langchain/core/runnables").RunnableConfig>;
-export { makeAssistant, Assistant };

package/dist/machine/machinev0.d.ts

@@ -1,163 +0,0 @@
-declare const machine: import("xstate").StateMachine<{
-    firstMessage: string;
-    voice: import("@fonoster/voice").VoiceResponse;
-    assistant: import("../assistants/assistants").Assistant;
-    playbackRef: string;
-    speechBuffer: string;
-    speechResponseStartTime: number;
-    speechResponseTime: number;
-}, {
-    type: "SPEECH_START";
-} | {
-    type: "SPEECH_END";
-} | {
-    type: "SESSION_END";
-} | {
-    type: "HUMAN_PROMPT";
-    speech: string;
-}, {}, never, import("xstate").Values<{
-    appendSpeech: {
-        type: "appendSpeech";
-        params: unknown;
-    };
-    hangup: {
-        type: "hangup";
-        params: unknown;
-    };
-    sendGreeting: {
-        type: "sendGreeting";
-        params: unknown;
-    };
-    interruptMachineSpeaking: {
-        type: "interruptMachineSpeaking";
-        params: unknown;
-    };
-    processHumanRequest: {
-        type: "processHumanRequest";
-        params: unknown;
-    };
-}>, {
-    type: "hasSpeechBuffer";
-    params: unknown;
-}, never, "hangup" | "welcome" | "machineListening" | "humanSpeaking", string, {
-    firstMessage: string;
-    voice: import("@fonoster/voice").VoiceResponse;
-    assistant: import("../assistants/assistants").Assistant;
-}, import("xstate").NonReducibleUnknown, import("xstate").EventObject, import("xstate").MetaObject, {
-    readonly context: ({ input }: {
-        spawn: {
-            <TSrc extends never>(logic: TSrc, ...[options]: never): import("xstate").ActorRefFromLogic<never>;
-            <TLogic extends import("xstate").AnyActorLogic>(src: TLogic, options?: {
-                id?: never;
-                systemId?: string;
-                input?: import("xstate").InputFrom<TLogic> | undefined;
-                syncSnapshot?: boolean;
-            } | undefined): import("xstate").ActorRefFromLogic<TLogic>;
-        };
-        input: {
-            firstMessage: string;
-            voice: import("@fonoster/voice").VoiceResponse;
-            assistant: import("../assistants/assistants").Assistant;
-        };
-        self: import("xstate").ActorRef<import("xstate").MachineSnapshot<{
-            firstMessage: string;
-            voice: import("@fonoster/voice").VoiceResponse;
-            assistant: import("../assistants/assistants").Assistant;
-            playbackRef: string;
-            speechBuffer: string;
-            speechResponseStartTime: number;
-            speechResponseTime: number;
-        }, {
-            type: "SPEECH_START";
-        } | {
-            type: "SPEECH_END";
-        } | {
-            type: "SESSION_END";
-        } | {
-            type: "HUMAN_PROMPT";
-            speech: string;
-        }, Record<string, import("xstate").AnyActorRef | undefined>, import("xstate").StateValue, string, unknown, any, any>, {
-            type: "SPEECH_START";
-        } | {
-            type: "SPEECH_END";
-        } | {
-            type: "SESSION_END";
-        } | {
-            type: "HUMAN_PROMPT";
-            speech: string;
-        }, import("xstate").AnyEventObject>;
-    }) => {
-        firstMessage: string;
-        voice: import("@fonoster/voice").VoiceResponse;
-        assistant: import("@langchain/core/runnables").Runnable<any, string, import("@langchain/core/runnables").RunnableConfig>;
-        playbackRef: string;
-        speechBuffer: string;
-        speechResponseStartTime: number;
-        speechResponseTime: number;
-    };
-    readonly id: "fnAI_v0";
-    readonly initial: "welcome";
-    readonly states: {
-        readonly welcome: {
-            readonly entry: {
-                readonly type: "sendGreeting";
-            };
-            readonly always: {
-                readonly target: "machineListening";
-            };
-            readonly description: "The initial state where the AI greets the Human.";
-        };
-        readonly machineListening: {
-            readonly on: {
-                readonly SPEECH_START: {
-                    readonly target: "humanSpeaking";
-                    readonly description: "This must be triggered by a VAD or similar system.";
-                };
-                readonly HUMAN_PROMPT: {
-                    readonly actions: {
-                        readonly type: "appendSpeech";
-                    };
-                    readonly description: "Appends the speech to the buffer.";
-                };
-            };
-            readonly description: "The state where the AI is actively listening in conversation.";
-        };
-        readonly humanSpeaking: {
-            readonly entry: {
-                readonly type: "interruptMachineSpeaking";
-            };
-            readonly on: {
-                readonly HUMAN_PROMPT: {
-                    readonly actions: {
-                        readonly type: "appendSpeech";
-                    };
-                    readonly description: "Appends the speech to the buffer.";
-                };
-                readonly SPEECH_END: {
-                    readonly target: "machineListening";
-                    readonly actions: {
-                        readonly type: "processHumanRequest";
-                    };
-                    readonly guard: {
-                        readonly type: "hasSpeechBuffer";
-                    };
-                    readonly description: "This must be triggered by a VAD or similar system.";
-                };
-            };
-            readonly description: "The state where the AI detects Human speech while it is speaking.";
-        };
-        readonly hangup: {
-            readonly type: "final";
-            readonly entry: {
-                readonly type: "hangup";
-            };
-            readonly on: {
-                readonly SESSION_END: {
-                    readonly target: "hangup";
-                };
-            };
-            readonly description: "The final state where the AI terminates the conversation due to inactivity.";
-        };
-    };
-}>;
-export { machine };

package/dist/machine/machinev0.js

@@ -1,142 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.machine = void 0;
-/*
- * Copyright (C) 2024 by Fonoster Inc (https://fonoster.com)
- * http://github.com/fonoster/fonoster
- *
- * This file is part of Fonoster
- *
- * Licensed under the MIT License (the "License");
- * you may not use this file except in compliance with
- * the License. You may obtain a copy of the License at
- *
- *    https://opensource.org/licenses/MIT
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-const common_1 = require("@fonoster/common");
-const logger_1 = require("@fonoster/logger");
-const uuid_1 = require("uuid");
-const xstate_1 = require("xstate");
-const types_1 = require("./types");
-const logger = (0, logger_1.getLogger)({ service: "autopilot", filePath: __filename });
-const machine = (0, xstate_1.setup)({
-    types: types_1.types,
-    actions: {
-        sendGreeting: async function ({ context }) {
-            await context.voice.answer();
-            await context.voice.say(context.firstMessage, {
-                playbackRef: context.playbackRef
-            });
-        },
-        interruptMachineSpeaking: async function ({ context }) {
-            logger.verbose("interrupting the machine", {
-                playbackRef: context.playbackRef
-            });
-            await context.voice.playbackControl(context.playbackRef, common_1.PlaybackControlAction.STOP);
-        },
-        appendSpeech: function ({ context, event }) {
-            const speech = event.speech;
-            context.speechBuffer = (context.speechBuffer || "") + " " + speech;
-            context.speechResponseStartTime = Date.now();
-            logger.verbose("appended speech", { speechBuffer: context.speechBuffer });
-        },
-        processHumanRequest: async function ({ context }) {
-            const speech = context.speechBuffer.trim();
-            logger.verbose("processing human request", { speech });
-            const response = await context.assistant.invoke({
-                text: speech
-            });
-            const speechResponseTime = Date.now() - context.speechResponseStartTime;
-            context.speechResponseTime = speechResponseTime;
-            logger.verbose("assistant response", {
-                response,
-                responseTime: speechResponseTime
-            });
-            await context.voice.say(response, { playbackRef: context.playbackRef });
-            // Clear the speech buffer and reset response timing
-            context.speechBuffer = "";
-            context.speechResponseStartTime = 0;
-        },
-        hangup: async function ({ context }) {
-            await context.voice.hangup();
-        }
-    },
-    guards: {
-        hasSpeechBuffer: function ({ context }) {
-            return context.speechBuffer?.trim().length > 0;
-        }
-    }
-}).createMachine({
-    context: ({ input }) => ({
-        firstMessage: input.firstMessage,
-        voice: input.voice,
-        assistant: input.assistant,
-        playbackRef: (0, uuid_1.v4)(),
-        speechBuffer: "",
-        speechResponseStartTime: 0,
-        speechResponseTime: 0
-    }),
-    id: "fnAI_v0",
-    initial: "welcome",
-    states: {
-        welcome: {
-            entry: {
-                type: "sendGreeting"
-            },
-            always: {
-                target: "machineListening"
-            },
-            description: "The initial state where the AI greets the Human."
-        },
-        machineListening: {
-            on: {
-                SPEECH_START: {
-                    target: "humanSpeaking",
-                    description: "This must be triggered by a VAD or similar system."
-                },
-                HUMAN_PROMPT: {
-                    actions: { type: "appendSpeech" },
-                    description: "Appends the speech to the buffer."
-                }
-            },
-            description: "The state where the AI is actively listening in conversation."
-        },
-        humanSpeaking: {
-            entry: {
-                type: "interruptMachineSpeaking"
-            },
-            on: {
-                HUMAN_PROMPT: {
-                    actions: { type: "appendSpeech" },
-                    description: "Appends the speech to the buffer."
-                },
-                SPEECH_END: {
-                    target: "machineListening",
-                    actions: { type: "processHumanRequest" },
-                    guard: { type: "hasSpeechBuffer" },
-                    description: "This must be triggered by a VAD or similar system."
-                }
-            },
-            description: "The state where the AI detects Human speech while it is speaking."
-        },
-        hangup: {
-            type: "final",
-            entry: {
-                type: "hangup"
-            },
-            on: {
-                SESSION_END: {
-                    target: "hangup"
-                }
-            },
-            description: "The final state where the AI terminates the conversation due to inactivity."
-        }
-    }
-});
-exports.machine = machine;

package/dist/machine/types.d.ts

@@ -1,29 +0,0 @@
-import { VoiceResponse } from "@fonoster/voice";
-import { Assistant } from "../assistants/assistants";
-declare const types: {
-    context: {
-        firstMessage: string;
-        voice: VoiceResponse;
-        assistant: Assistant;
-        playbackRef: string;
-        speechBuffer: string;
-        speechResponseStartTime: number;
-        speechResponseTime: number;
-    };
-    input: {
-        firstMessage: string;
-        voice: VoiceResponse;
-        assistant: Assistant;
-    };
-    events: {
-        type: "SPEECH_START";
-    } | {
-        type: "SPEECH_END";
-    } | {
-        type: "SESSION_END";
-    } | {
-        type: "HUMAN_PROMPT";
-        speech: string;
-    };
-};
-export { types };

package/dist/machine/types.js

@@ -1,9 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.types = void 0;
-const types = {
-    context: {},
-    input: {},
-    events: {}
-};
-exports.types = types;

package/dist/runner.d.ts

@@ -1,2 +0,0 @@
-#!/usr/bin/env node
-export {};

package/dist/vad/micVadTest.js

@@ -1,48 +0,0 @@
-"use strict";
-var __importDefault = (this && this.__importDefault) || function (mod) {
-    return (mod && mod.__esModule) ? mod : { "default": mod };
-};
-Object.defineProperty(exports, "__esModule", { value: true });
-/*
- * Copyright (C) 2024 by Fonoster Inc (https://fonoster.com)
- * http://github.com/fonoster/fonoster
- *
- * This file is part of Fonoster
- *
- * Licensed under the MIT License (the "License");
- * you may not use this file except in compliance with
- * the License. You may obtain a copy of the License at
- *
- *    https://opensource.org/licenses/MIT
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-const logger_1 = require("@fonoster/logger");
-const node_record_lpcm16_1 = __importDefault(require("node-record-lpcm16"));
-const makeVad_1 = require("./makeVad");
-const logger = (0, logger_1.getLogger)({ service: "autopilot", filePath: __filename });
-async function main() {
-    const vad = await (0, makeVad_1.makeVad)();
-    // Start recording from the default microphone
-    const mic = node_record_lpcm16_1.default
-        .record({
-        sampleRate: 16000, // 16 kHz sample rate
-        channels: 1,
-        threshold: 0.5
-    })
-        .stream();
-    mic.on("data", async (data) => {
-        const chunk = new Uint8Array(data.buffer);
-        await vad(chunk, (event, _data) => {
-            logger.info("vad event:", { event, data: _data });
-        });
-    });
-    mic.on("error", (err) => {
-        logger.error("an error occurred:", { err });
-    });
-}
-main().catch(logger.error);