You can programmatically access hack.chat using the following commands via a websocket. To prevent getting disconnected, ping the server every 50 seconds. A list of wrappers written for accessing hack.chat can be found [here](https://github.com/hack-chat/3rd-party-software-list#libraries).
The commands are to be sent through a websocket to the URL `wss://hack.chat/chat-ws` (everything sent and received are `string`s). If you are sending messages locally or to another domain, replace 'hack.chat' with the respective domain. If you're running your own instance of hack.chat, you can retain backwards-compatibility in order to ensure that software created for the main server will work on yours too.
All commands sent must be JSON objects with the command specified in the `"cmd"` key. For example:
hack.chat has three permission levels. When you access a command, hack.chat automatically knows your permission level from your trip code. The lowest permission level is `core`. `mod` is above `core`, so it can access `core` commands in addition to `mod` commands. `admin` is similarly above `mod`.
|`help`|`category` or `command`|Gives documentation programmatically. If `category` (the permission level, such as `mod`) is sent, a list of commands available to that permission level will be sent back (as a `string` and not an `array`). This list only includes what is unique to that category and not every command a user with that permission level could perform. If `command` (e.g., `chat`) is sent, a description of the command will be sent back.|