Round one, as designed
Ten pilot projects across DeFi (3), NFT/gaming (2), AI (2), tooling (2), identity (1). Each commits to ≥ 100 verifiable testnet transactions, open source, a demo, a reproducible README and at least one native pallet. Support in testnet DMS, core-team access, explorer listing, public mention. Status: not open — faucet, explorer and templates pending.
What round two would require
Completed projects from round one. The grants document's table has ten rows; a second round makes sense when most are marked completed — meaning a stranger verified the five KPIs — not when the calendar says a quarter has passed.
A network that absorbed them. Ten projects each sending 100+ transactions is the first sustained external load the testnet will have seen. If the public RPC's rate limit, the two-validator finality or the explorer's indexer buckled, round two waits for the fix.
Gaps closed. Round one's most valuable output will be the list of things that did not work — pallets that reverted, docs that misled, SDK calls that failed. Round two opens after that list is worked, so the second cohort does not hit the first cohort's bugs.
Decentralization progress. If round one produced no external validator operators, the programme's tooling category should be re-weighted toward node operation.
Quotas re-balanced by evidence. The committee may adjust the five categories' slots based on the quality of round-one proposals, keeping the total at ten.
What would not trigger round two
Demand. A hundred applications for round one is a reason to be selective, not to double the programme before the network has been tested by the first ten.
The larger sequence
Faucet → explorer → templates → hackathon → grants round one → findings worked → round two. Each step is gated on the previous one being real. The page this replaces announced a second wave of houses on the strength of the first; this ecosystem's second anything is announced on the strength of a table with checkmarks in it.
Based on TESTNET_GRANTS_PROGRAM.md §2, §5 and §8.