Clientv2 is an example client for Watson Assistant (WA) applications. Its purpose is to test and demonstrate usage of dialog API, demonstrate client side functionality (e.g. providing information using the external services (jokes, news) or controlling custom made hardware), and to act as fast prototyping customizable front-end for accessing dialog applications.
This project is in active development. Do not hesitate to contact us. Any feedback is appreciated.
- Go to config tab
- Fill in your conversation service credentials.
- Click reload
- Select workspace
See Configuration for more details.
Example configured link
This link caries example service credentials that are passed to config tab.
- if you get stuck at "no privileged call" page while logging in to IBM Cloud try again in incognito mode (ctrl + shift + n).
This app contains following functionality:
- Chat window
- Text formatting with html
- Renders buttons and other arbitrary html
- Delayed messages and reminders
- Input can be specified as text or json directly into input box
- IBM Cloud speech recognition is integrated as alternative input. It starts by pressing microphone button next to input box.
- Alternative output methods
- Compact layout mode suitable for presentation and embedding in web pages. See Embedded chat window example.
- Adding external entities is supported (currently from AlchemyAPI)
- Plugable modules for the services (implemented jokes, news, weather, stock market, namedays, firmanta protocol)
- Multi workspace setup is supported (e.g. application + chitchat)
- One can connect simple hardware, move motors, blink LEDs or control robots (for version running locally only)
- Developer tools
- bare json input/reply panels, log
- config tab for customization
Instalation of local instance
- Install node.js from https://nodejs.org/en/download/
- Fork repository https://hub.jazz.net/project/tmacek/clientv2t/ or unpack zip the folder of a choice
npm install (will install all the libraries)
node app.js (runs the application)
- connect in browser to http://localhost:3000/
To get more information about how to use the app see Highlights.
See full Documentation for more insight into how the clientv2 works.