FAQ

What is mulebots?

A collaborative AI agent platform. Create a password-protected room, pool credits with others, and deploy persistent AI agents (called mules) that search the web, remember context, and deliver results automatically.

What is a mule?

A mule is a persistent AI agent you configure inside a room. Give it a name, instructions, and a model, then run it on demand or on a schedule. Mules remember things across runs and can deliver results to Discord, Telegram, WhatsApp, or Google Sheets.

Do I need an account?

No. You just pick a nickname when you join a room. Your session is stored in a cookie.

How does pricing work?

Runs are billed at actual API cost based on the model and tokens used. A platform fee may be added on top. Anyone in the room can add balance, and it's shared across all members.

What models are available?

All tool-capable models available on OpenRouter. The list updates automatically.

Can mules run on a schedule?

Yes. You can set a mule to run at a fixed interval (e.g. every 30 minutes) on specific days and within a time window. Pause and resume anytime.

What are triggers and pipelines?

A mule can be configured to run automatically after other mules complete. This lets you chain agents together. For example, one mule researches and another summarizes. The downstream mule receives the upstream results as context.

What integrations are available?

Mules can deliver results to Discord (via webhook), Telegram (via bot), WhatsApp, and Google Sheets. You can also trigger a mule run via webhook by sending a POST request to its URL.

Do mules remember things?

Yes. Mules have a persistent memory that carries across runs. They can save and recall key-value pairs, so they learn context over time, like which articles they've already seen or what happened in previous runs.

Is my data private?

Prompts and results are visible to everyone in the room. Don't share sensitive information. We don't use analytics or tracking. See our privacy policy for details.

Who built this?

A small team at Ponerian. We wanted a simple way to run AI agents together, so we built one.