Sigma Maze Labs · Hamburg
Sigma Maze Labs builds products at the intersection of AI and blockchain: machine judgement on on-chain data, shipped, logged and measured against what actually happened.
Thesis
Every launch, trade and withdrawal is public and permanent. The data is there; the judgement isn't.
Raw events don't help anyone. A scored, explained decision ("flagged as liquidity pull, 0.91, here's the tx") does.
Each prediction is logged with an id and later joined with the real outcome. Our models get better from evidence, not opinion.
Products
Live rug-pull detection for Solana and Robinhood Chain, with the on-chain evidence for every flag.
Push alerts for mosh.trade bundles on Robinhood Chain: new raises, graduations, and every move of the agent vaults.
Independent, side-by-side ranking of hardware crypto wallets: specs, prices, open-source status.
Production-grade validator infrastructure for public blockchains since 2018.
Our products are automated, probabilistic analytics. Nothing on this site is financial advice.
Method
In the lab
What we're exploring next. No dates, no promises.
Paste a token, get the odds it rugs, before you buy.
RugPoolRadar and bundle signals as an API and MCP server, so AI agents can ask "is this token safe?" before they trade.
The same observe-score-learn loop, beyond Solana and Robinhood Chain.
Principles
Every claim links to a transaction.
We say "flagged as" and show confidence. We never name people.
If a decision isn't logged, it didn't happen.
This site sets no cookies and loads nothing from third parties.
Partnerships, data, sponsorship, or you want to build with us.
