Channels you can test on

The two channels a test can run on today, web chat or API and Telegram, and what each one needs from you.

A channel is the way our attacker reaches your agent. Two channels run today. The console offers a channel only when it is in your plan and a test can run on it.

Channel	Shown in the console as	What you give us	How you prove it is yours
Web chat or API	Web chat or API	A public https address your agent answers on	A text file on the same site
Telegram	Telegram	Your bot’s user name, such as @AcmeHelpBot	A code in the bot’s About text

​

Web chat or API

We send each attacker message to your address as a signed POST. Your agent answers with JSON. One conversation keeps the same conversation id, so your agent can keep context.

What the address must be:

Rule	Detail
Scheme	https only
Port	443 only
Host	A public name with a dot. Not localhost, not .local, not .internal, not an IP address on a private network
Credentials	None in the address
DNS	Every address the name points to must be public. We check this when you save, when you prove it and before every test message

See Connect a web agent for the request, the reply and the signature.

​

Telegram

Your agent is a Telegram bot. Our attacker writes to it as a Telegram user account, because on Telegram a bot cannot start a chat with another bot. The attacker sends the text as written and reads the bot’s answer.

Rule	Detail
Address	The bot’s user name, 5 to 32 letters, digits or underscores, ending in “bot”
Reply wait	Up to 4 minutes for an answer
Reply length	The first 16,000 characters are kept

See Connect a Telegram bot.

​

Channels in your plan that are not ready

Your plan may list more channels than these two. The add form shows them under “Also in your plan, not ready for testing yet”. You cannot add an agent on them. See Plans and usage.

​

One channel per agent

An agent you add has one channel and one address. To test the same agent on another channel, add it again with the other address.