Twiliomatic

A TwiML generator using GPT-3 from OpenAI

Setup

• Install Twilio CLI

brew tap twilio/brew && brew install twilio

• Install Twilio Serverless Plugin

twilio plugins:install @twilio-labs/plugin-serverless

Setup your environment

From the root of your repository

npm install
npx configure-env

Develop locally

twilio serverless:start

Deploy to dev

twilio serverless:deploy

Deploy to prod

twilio serverless:deploy --production

The main interface currently lives at /index

Use the output from your deploys to locate the full url.

It will look something like:

https://twiliomatic-9999-dev.twil.io/index

Where 9999 is unique to your deploy and environment.

Scripts

change-mode

Provides the ability to change modes by environment. keynote for when it airs. shutdown to turn things off.

npm run change-mode --mode=keynote --env=production

npm run change-mode --mode=shutdown --env=dev

set-favorites

Update the shutdown mode favorites. Value is comma separated

npm run set-favorites --gists=ABC123,DEF456 --env=production

delete-all-gists

Removes all gists for the GITHUB_USER

npm run delete-all-gists

purge-cache

Purges cache for Cloudflare Zone CLOUDFLARE_ZONE_ID

npm run purge-cache

Overview

The OpenAI prompts are defined in twiml-prompts

/mode

Returns an object that helps to detect the current state of things

{
  "mode": "keynote",
  "enabled": true
}

/requester

Allows for a long running query through the use of Sync.

Parameter	Description
Query	(Required intiially) The prompt you'd like to complete
ID	The ID of the request delivered after the initial Query is ran

See client side long-query.js and it's usage

Tests

There are not too many 😢

npm test