Skip to main content

Bot

Open platform bot are interactive robot solutions

Main process below:

  • Create openapp
  • Get appid and appsecret
  • Enter the bot page and enable the bot ability
  • Fill in the callback address with the public http service address that can be accessed
  • Go back to the group page, enter the appid in the group details integration function, find the app you just created, and add it to the group
  • In the group @bot and enter text, tailchat will send an http request with message content to the address shown in the callback address
  • Receive the request sent by tailchat in the bot service, and send a response to the corresponding panel of the corresponding group through the bot

Of course, create in Tailchat before you start anything

Let's take a look at the actual development process of the bot:

Development with SDK (Node.js)​

Tailchat provides sdk as a rapid development tool, tailchat-client-sdk, you can install it with the following command

npm install tailchat-client-sdk

Taking koa as an example, we first create a simple koa service as follows:

Create a node project:

mkdir tailchat-bot && cd tailchat-bot
npm init -y
npm install koa koa-router tailchat-client-sdk

Create server.js file:

const Koa = require('koa');
const Router = require('koa-router');
const app = new Koa();
const router = new Router();

// Define route
router.get('/', async (ctx) => {
  ctx.body = 'Hello, World!';
});

router.post('/bot/callback', async (ctx) => {
  ctx.body = 'Bot Callback Page';
});

// Register middleware
app.use(router.routes());
app.use(router.allowedMethods());

// Start server
app.listen(3000, () => {
  console.log('Server is running on http://localhost:3000');
});

At this point we have established two basic routes, / and /bot/callback, and listened to 3000 port, note that /bot/callback listens to POST requests

At this point we execute node server.js to see that our application will be started.

Now we need to add some logic, for example, if we want to implement a repeating bot, then modify the implementation of /bot/callback route as follows:

import { TailchatHTTPClient, stripMentionTag } from 'tailchat-client-sdk';

const host = '<your tailchat instance backend host>';
const appId = '<appId>';
const appSecret = '<appSecret>';

const client = new TailchatHTTPClient(host, appId, appSecret)

// ...

router.post('/bot/callback', async (ctx) => {
  const type = ctx.body.type;
  
  if (type === 'message') {
    const payload = ctx.body.payload;
    try {
      const message = await client.replyMessage({
        messageId: payload.messageId,
        author: payload.messageAuthor,
        content: payload.messageSnippet
      }, {
        groupId: payload.groupId,
        converseId: payload.converseId,
        content: `Your message: ${stripMentionTag(payload.messageSnippet)}`,
      })

      console.log('send message success:', message)
    } catch (err) {
      console.log('send message failed:', err)
    }
  }
  
  ctx.body = 'Bot Callback Page';
});

Please fill in host, appId and appSecret into the appId and appSecret obtained during creation, and host into the address of the Tailchat server, the official address of nightly is https //tailchat-nightly.moonrailgun.com

The content of the reply is not important, just make sure not to actively return an error message, Tailchat does not care about the returned content

Please note that if you want to share your code, please keep your appSecret, which is equivalent to your account password

The logic is very simple. Get the message content, author, id, group id, and converse id from the request. Send content as a reply

Deploy the application online to see the effect.

info

Before testing, please make sure you have enabled the bot ability and filled in the correct callback address

Develop in other languages​

Since it is a network application, it is of course not limited to nodejs. The following are the formats of some network requests that need to be used, mainly sending requests to the Tailchat server as an open platform bot.

The official nightly api address is https://tailchat-nightly.moonrailgun.com, please replace it with your own backend address for self-deployment

Login​

Before all requests, you need to log in to obtain the jwt token to indicate your identity, and you need to send the following content:

POST /api/openapi/bot/login

Header
Content-Type: application/json

Body
{
  appId: <your app id>,
  token: <md5(appId+appSecret)>,
}

Response
{
  data: {
    jwt: "..........",
    userId: ".........."
  }
}

The token of the request body is a fixed value, which needs to be encrypted with the md5 algorithm after splicing appId and appSecret. Finally get jwt, jwt must be on the request header in all subsequent requests

Header
X-Token: <your-jwt>

Call​

Robots can call the interface like ordinary users, such as:

POST /api/xxx

Header
Content-Type: application/json
X-Token: <your-jwt>

Body
{
  ...
}

You can treat the bot as an ordinary user. The bot can do everything that ordinary users can do, and the bot will also be subject to the permission restrictions that ordinary users need to be subject to.

The difference is that regular users interact with visualizations, while bots interact with APIs.

Send Message​

POST /api/chat/message/sendMessage

Header
Content-Type: application/json
X-Token: <your-jwt>

Body
{
  "converseId": "",
  "groupId": "",
  "content": "",
  "plain": "",
  "meta": {},
}

Reply message​

POST /api/chat/message/sendMessage

Header
Content-Type: application/json
X-Token: <your-jwt>

Body
{
  "converseId": "<converId/panelId>",
  "groupId": "<groupId, optional in DM>",
  "content": "<your message content>",
  "plain": "<your plained message, optional>",
  "meta": {
    mentions: ["<replyMessageAuthorId>"],
    reply: {
      _id: "<replyMessageId>",
      author: "<replyMessageAuthor>",
      content: "<replyMessageContent>",
    },
  },
}

Additional Documentation​