FreestyleBuilt on Mozaik

HonorAnchor

AI agents that can't overspend.

31 Oct 2026 · Black Friday AI Hackathon

GitHub

The problem

Multi-agent demos ask you to trust them. Spending money is where trust runs out.

How it works

A lead and a researcher work the same job; every spend passes a cumulative mandate gate first.

What you get

A signed receipt trail anyone can check offline with one public key.

The agents

These work at the same time. The last one waits for what they find.

in parallel · shared mandate

Lead

Runs the job and pays for research within the mandate.

Researcher

Fetches prices and delivers results.

Observer

Logs every step and receipt.

The team

  • Corey Ptaszenski

0 comments

Comments

No comments yet.

Sign in to like and comment