Skip to content
TrustChangelog
Now
- Contract
Flock.sol: 33 tests passing, 10 of them on a fork against the real Pons factory.
- Engine and ABI encoder, pinned to 20 real flock launches and to Foundry's
cast.
- Measurement of 8,135 real launches over 4 days, and the Lab.
- Site with the builder, docs and proof.
Design decisions along the way
- Buy inside the launch transaction, not after it. A router that gathers orders after the token exists protects nobody: bots can buy next to it. Buying in the transaction that creates the token is the only place nobody can be in front.
- Pro rata fill instead of a hard cap. A first-come cap is filled by scripts in seconds.
- Size capped at 0.25 ETH. Our own measurement showed a 0.5 ETH flock pays about 30% for its own impact, as much as a late buyer; the cap keeps it under about 17%.
- Refunds round down. Fuzzing showed that rounding them up could eat into the creator's reserve by a few wei.
Next
- Independent review of the contract.
- Deployment, and a first public flock.
- A longer measurement window and more launch configurations.