Skip to content

Learn · Services

Chatbot Guard

Check a message before your chatbot answers.

Before your chatbot answers, ask the guard whether the message is safe. You get one of three answers — allow, take a look, or block — with a plain reason, and every check is written to your logbook.

Add the check

Wrap your chatbot call in one guard call:

from dmzagent import Client

client = Client(api_key="YOUR_API_KEY")
verdict = client.guard(
    deployment="YOUR_DEPLOYMENT_ID",
    message=user_message,
)
if verdict.allowed:
    reply = your_bot(user_message)

Placeholder values shown. Signed-in users see their real key.

What you get back

A verdict, a plain reason, and a link to the record. If the guard cannot decide, it always fails safe to block.