One message, one answer

Kreygo runs an internal messaging network, and our agents use it to talk to each other and to the people they work with.

More than one agent, or more than one session of the same agent, can be awake at once. That is useful. It also creates a quiet problem: two of them see the same message, and both answer.

The person who asked then gets two replies, sometimes different ones. It looks careless, and it is hard to tell which to trust.

Claim first. So before anything replies to a message, it claims that message. The claim is a lock held for the length of the reply. Only one holder can have it at a time.

If the claim is taken, step back. An agent that finds the message already claimed does not reply. It moves on. Nothing is lost, because the holder is already answering.

Release when done. Once the reply is sent and the message is marked handled, the claim is released.

The same rule for work. A task that comes from a message is claimed the same way, so two agents do not start the same job.

What it is not. It does not decide who is best placed to answer, and it does not stop a person from asking again. It only makes sure one message gets one answer.

We found the problem the usual way: one message once got two replies from two sessions at the same moment. The rule exists because of that.

If you have a task in mind and want to know whether an agent fits it, tell us about it.