DEPLOY_EN.md

Deployment Process

Manual Deployment

Step One: Create a Telegram Bot and Obtain a Token

1. Open Telegram and send the /start command to BotFather.
2. Send the /newbot command to BotFather and give your bot a name.
3. Give your bot a unique username ending with _bot.
4. BotFather will generate a token. Copy and save it. This token is the key that binds the bot to your account. Do not disclose it to others!
5. Later, in the Cloudflare Workers settings, fill in the TELEGRAM_TOKEN variable with this token.

Step Two: Register with OpenAI and Create an API Key

1. Open OpenAI to register an account.
2. Click on the avatar in the upper right corner to enter the personal settings page.
3. Click on API Keys and create a new API Key.
4. Later, in the Cloudflare Workers settings, fill in the API_KEY variable with this key.

Step Three: Deploy Workers

1. Open Cloudflare Workers to register an account.
2. Click on Create a Service in the upper right corner.
3. Enter the newly created workers, select Quick Edit, copy the ./dist/index.js code into the editor, and save.

Step Four: Configure Environment Variables

1. Open Cloudflare Workers and click on your workers. Click on Setting -> Variables in the upper right corner.
2. API_KEY: Set it to the OpenAI API Key.
3. TELEGRAM_AVAILABLE_TOKENS: Set it to the Telegram Bot Token.
4. WORKERS_DOMAIN: Set it to your Workers domain, for example, workers_name.username.workers.dev. You can find it in the Preview of the workers details page.
5. CHAT_WHITE_LIST: Set it to the ID of the users who are allowed to access, for example, 123456789,987654321. If you don't know your ID, use the /new command in a conversation with the bot to get it.
6. I_AM_A_GENEROUS_PERSON: If you still don't understand how to get the ID, you can set this value to true to turn off the whitelist function and allow everyone to access.

Step Five: Bind KV Data

1. On the Home-Workers-KV page, click Create a Namespace in the upper right corner. You can name it whatever you like, but when binding, it must be set to DATABASE.
   
2. Open Cloudflare Workers and click on your workers.
3. Click on Setting -> Variables in the upper right corner.
   
4. Click Edit variables in KV Namespace Bindings.
5. Click Add variable.
6. Set the name to DATABASE and select the KV data you just created.

Step Six: Initialization

1. Run https://workers_name.username.workers.dev/init， Automatically bind telegram's webhook and set all commands.

Step Seven. Starting a Chat

1. To start a new conversation, use the /new command. The chat context will then be sent to ChatGPT every time.
2. Use the /setenv KEY=VALUE command to modify user settings, such as SETENV SYSTEM_INIT_MESSAGE="Starting now, every sentence will end with 'meow'."
3. Because every conversation includes all the historical records and may reach the 4096-token limit, it is recommended to clean up the history by using the /new command when there is nothing to discuss.

Automatic Deployment

0. Manual deployment steps one, two, three
1. Modify the configuration file wrangler-example.toml to wrangler.toml
2. npm install
3. npm run login
4. npm run deploy