A custom Hasura service that acts as an Intercom Webhook. The project allows one to create a very basic static keyword - response type bot. The frontend interface has an editable table that one can use to add the keyword - response pairs. Testing the bot is easily done with the Intercom chat app on the bottom right section of the screen.
Hasura hub: https://hasura.io/hub/project/omnicron/intercom-webhook
hasura cluster add <cluster-name> <alias> (Not required if cloned through hasura quickstart)
hasura cluster set-default <alias> (Not required if cloned through hasura quickstart)
To deploy the project in your own cluster, all you need to do is set the proper cluster name value in the ui microservice to that of yours.
- Replace the cluster name in
microservices/ui/app/src/components/Config.js - Replace the cluster name in
microservices/ui/app/package.json, in thestartandbuildkeys of thescriptsobject.
When done, deploy the code to your cluster using git push <alias> master.
The project contains two microservices, ui for the frontend, in ReactJS and api for the backend in NodeJS.
The frontend is completely built using custom components, while using react-router for SPA redirection and react-motion for transition animations.
Every major component is written within a separate file to making them easier to reuse and manage. The style for each such component is also available as ui/app/src/components/css/<Component>.css, where Component is the filename of the corresponding React component.
Most of the backend code is present in the microservices/api/src/server.js file.
The REST API involves the following endpoints:
/intercom(GET): A dummy endpoint to test if the server is working./intercom_webhook(POST): Processes and replies to webhook notifications from Intercom. This is where the app generated replies to the user are dispatched from./add_entries(POST): Accepts JSON object with the format{ data: [Object] }, wheredatais an array of objects of type{ key: <string>, message: <string> }. Keys are case-insensitive. This submitted data is stored in the database which can then be used in the Intercom bot./get_value(GET): Endpoint to test the return message for a given key. The format for the request could look likehttps://api.cluster-name.hasura-app.io/get_value?key=hello.