A trustless FIL to WFIL bridge, and the liquidity bound at its centre

Wrapped Filecoin had launched on Ethereum through Anchorage and Tokensoft: an ERC-20 backed one-to-one by FIL held in custody by a US trust. That works, and it makes every wrap a permissioned action through a regulated custodian. FIL holders wanting DeFi on Ethereum or Polygon either accepted that or did something worse, which in practice meant selling FIL for a stablecoin and buying WFIL on a DEX, paying spread twice and taking price risk in between.

The technical substance is what replaces the custodian, because the custodian was not merely holding coins, it was the guarantee that the wrapped supply is backed. Removing it means backing has to be provable from chain state rather than asserted by an auditor, across two chains with very different finality models.

The proposal names its binding constraint correctly, which most bridge pitches do not: insufficient liquidity does not put a holder’s funds at risk, it limits how much can be bridged at once. That is a size bound, not a safety risk, and conflating the two is how bridges end up with security arguments that are really liquidity arguments.

Submitted to filecoin-project/devgrants. Sixty days to launch, one week for the formal spec. That estimate was optimistic, and this is the closest thing here to a design document written before building rather than after.