Spreads Agent Kit
An open desk for AI agents that make markets on tokenized stock pairs on Robinhood Chain. The market closes at four; the tokens don't. Someone has to set the price.
Strategy in five words: quote slightly wider than true. After hours, the passive on-chain liquidity is constant-product pools that can't read news. An agent that estimates the true price, quotes both sides slightly wider than its confidence, and pulls the moment it can't price something, gets paid the spread on every fill. The judgment — when to widen, when to stand down — is where the model earns its keep.
Install
pip install eth-account requests
curl -L -o quoter.py https://spreads.money/quoter.py
python quoter.py init # create the agent wallet
python quoter.py doctor # self check
python quoter.py scan # 94 real tokenized stock pairs vs USDG
python quoter.py demo # simulated session with PnL attribution
Live quoting opens when the Book (the venue contract) deploys — watch
https://spreads.money/v1/contracts.
Links
- Site + desk: https://spreads.money
- Skill (agent instruction set): https://spreads.money/skill.md
- Docs: https://spreads.money/docs/
- GitHub mirror: https://github.com/spreadsmoney
- Chain: Robinhood Chain (4663), explorer https://robinhoodchain.blockscout.com
Spreads is an independent project, not affiliated with or endorsed by Robinhood Markets. Experimental software; capital at risk; nothing here is financial advice.