/dialogflow-web

Web App for Dialogflow

Primary LanguageVueMIT LicenseMIT

Dialogflow for Web

Dialogflow lets you build conversational interfaces on top of your products and services by providing a powerful natural language understanding (NLU) engine to process and understand natural language input

This is a unofficial Web client for Dialogflow, it was built, to support Google Assistant rich-responses and to make the most out of the Platform available to Web

You can buy me some snacks, to support the development

Or just hire me

About Dialogflow V2 support

This repository currently does only support V1 API, you can still use this client till October 2019, but, when the new client arrives, this repository will be archieved and the focus will move towards the new one

Dialogflow for Web V2 will be released in Q1 2019 in a separate repository. I have fully rewritten and redesigned the client from scratch for even better experience. You will also need to install "Dialogflow Gateway" - a software, which connects Dialogflow gRPC API to HTTP Clients. It will also be released around the same date

What's included

  • 100% Progressive Web App, following Google's Material guidlines
  • Customizable environment
  • Works offline and without JavaScript
  • Responsive design
  • Contains Search Engine Optimization
  • Based on VueJS with Webpack & Live Reload
  • WebSpeech API for voice input
  • Text-To-Speech output
  • Perfect accessibility for disabled users

Requirements

  • NodeJS
  • NPM or Yarn
  • Basic knowledge in ES6 (JavaScript)
  • Dialogflow API Key (You can get it here)

What is supported

  • Basic reply (is being displayed and spoken)
  • Simple Response (is being spoken)
  • Cards
  • Carousel Cards
  • Lists
  • Suggestions Chips
  • Link chips
  • Combination of all above
  • Fancy animations and visual feedback
  • Widget mode, for more info see https://github.com/MishUshakov/dialogflow-widget
  • Googling unknown requests

Get started

Clone the repository to your machine

You can use git or download it directly from github

Get the dependencies

Open your cloned folder. Then, using your favorite package manager get the dependencies

Example using npm

npm install

Example using yarn

yarn

Start development server & build

Open your cloned folder. Then, using your favorite package manager run dev command

Example using npm

npm run dev

Example using yarn

yarn dev

When you done these steps, your default browser should open and redirect you to localhost:8080. If the port 8080 is already in use, you can give an port argument to connect at specified port

Example using npm

npm run dev --port 9090

Example using yarn

yarn dev --port 9090

PLEASE DO NOT USE THE DEVELOPMENT SERVER FOR SERVING YOUR APP, DEVELOPMENT SERVER IS MEANT FOR THE DEVELOPMENT. INSTEAD PLEASE UPLOAD BUILD ARTIFACTS TO A STATIC WEB-SERVER

Building for production

Your app will be bundled to dist/build.js To build it you can use npm or yarn

Example using npm

npm run build

Example using yarn

yarn build

Customizing App & Connecting your agent

To connect your bot, you first need to create one. You can do it by visiting Dialogflow website. Then create your intents and add Google Assistant responses

Open config.js and change the token variable, to connect your bot. You can also translate the strings to your language

Example

export default {
    app: {
        token: "9d686a47b1de48bab431e94750d1cd87"
    }
}

if you want to change the accent color open src/App.sass and change the $color variable

Frequently Asked Questions

  • Q: I changed the token and the responses are the same as with previous token
  • A: Make sure you have cleaned your cache and rebuilt your app. In Safari go to "Develop" > "Empty Caches". In Chrome: "Developer Tools" > "Application" > "Clear storage" > "Clear site data"

Thank you, stay updated!