This guide provides documentation for Wickr IO Integrations. If you're using AWS Wickr, see [AWS Wickr Administration Guide](https://docs.aws.amazon.com/wickr/latest/adminguide/what-is-wickr.html).

# Integration setup
<a name="integration-setup"></a>

Complete the following procedure to setup a new custom Wickr IO integration.

1. Create a new directory and install the Wickr IO bot API with NPM.

   ```
   mkdir wickrio-emoji-bot
   cd wickrio-emoji-bot
   npm install wickrio-bot-api
   ```

1. In the root directory of your new package, create a file named `processes.json` with the following contents:

   ```
   {
     "apps": [
       {
         "name": "wickrio-emoji-bot",
         "env": {
           "tokens": {}
         }
       }
     ]
   }
   ```

   This file is used by the Wickr IO bot API library to provide custom configuration for your integration. Values provided in the `env.tokens` property will be set as environment variables in your integration.

1. In addition to the `processes.json` file, you need to create the following two scripts in order for your bot to start up successfully:

   1. **install.sh**

      This file is required when creating the bot. There are no special install time requirements, you can exit with a successful status (0).

      ```
      #!/bin/sh
      exit 0
      ```

   1. **start.sh**

      This script is used by the Wickr IO integration to start your bot. The code that is created below will be in the file `index.js`. Configure the startup script to run this file.

      ```
      #!/bin/bash
      node index.js "$@"
      ```

1. Make these scripts executable by running `chmod`:

   ```
   chmod +x {install,start}.sh
   ```

1. Next, create an `index.js` file which serves as the main entrypoint for our bot:

   ```
   // index.js
   const WickrIOBotAPI = require('wickrio-bot-api')
   
   const bot = new WickrIOBotAPI.WickrIOBot()
   const WickrIOAPI = bot.apiService().WickrIOAPI
   
   async function listen() {}
   
   async function main() {
     const username = process.argv[2]
     if (!username) throw new Error('Missing username')
   
     const status = await bot.start(username)
     if (!status) throw new Error('Unable to start bot')
   
     bot.startListening(listen);
   }
   
   main()
   ```