Coin Brief ENDE

A Kusama project sells one-cent queries over a group's knowledge graph, paid in dUSD

The Birdbrain group has opened what it calls the smallest working version of agents paying for what groups know, in a post on the Polkadot forum on 18 November. Its endpoint sells SQL queries over the group's knowledge graph - built from 42 weekly calls, with 713 nodes and 880 edges covering 29 sessions, plus 183 concepts from the recordings - for 0.01 dUSD per query plus 0.001 per row returned, up to 500 rows, on Kusama Asset Hub. The transcripts themselves are not for sale.

Payment needs no account and no API key. A buyer posts SQL and gets back an HTTP 402 response with an exact quote and free sample rows. The buyer transfers the quoted dUSD to Birdbrain's account, waits for finality, and sends the same request again with the block hash and an sr25519 or ed25519 signature over the SHA-256 hash of the SQL, which ties the transfer to that buyer and that query; each transfer can be claimed once. Asset Hub takes the fee in dUSD, so no KSM is needed, and settling costs 30 to 60 seconds.

The group says the endpoint has settled real payments on mainnet, but only from keys it holds, and is looking for a first outside buyer. It explains why it did not use x402 directly: its EVM schemes settle through EIP-3009 or Permit2 against an ERC-20 contract, native Asset Hub stablecoins are not EIP-3009 tokens, and an sr25519 account cannot use it. The design follows select402, a proof of concept by Parity's Karim Jedda that keeps the 402 response but lets the buyer pay with the chain's own keys.

A Kusama project sells one-cent queries over a group's knowledge graph, paid in dUSD
A Kusama project sells one-cent queries over a group's knowledge graph, paid in dUSD — Coin Brief

Why it matters

The interesting part is the proof, not the price: a signature over the exact query binds one payment to one answer without accounts or keys issued by the seller. Whether anyone outside the group pays is the test it has set itself.

Written by Victoria Shinder.