Alkanes Is the Limit of Bitcoin's L1. SUBFROST Is Its L0.

Bitcoin is one of the hardest protocols in existence to host an application on. For a developer looking to innovate, it is close to the most extreme choice available, and that is exactly what makes anything running on it credible.

Alkanes is notable because it is a pure Bitcoin stack. It relies on nothing but metashrew and access to a bitcoind for block synchronization (IBD); the whole of its logic lives on Bitcoin L1. Its state settles with every block, atomic with each Bitcoin transaction, and it extends Bitcoin without breaking a single rule. Bitcoin consensus natively supports something like Alkanes: application logic can coexist with money if you carry program deployments on ordinary transaction structures and encode protocol messages economically enough to stay cost-effective on-chain. The Protostone extension to the Runestone OP_RETURN structure does this well. Unlike an L2, Alkanes settles in a sovereign manner, with no sequencer and no rollup batching transactions on your behalf.

But there is a boundary L1 cannot cross on its own. Moving Bitcoin-native currency out, an exit, is not an application concern. It happens at the bridge layer. And for a Bitcoin-based protocol, that layer is L0.

Where L1 ends

On EVM chains, exiting an L2 is a solved shape: smart contracts, Merkle proofs, and zk verifiers settle the claim back on-chain. On Bitcoin, your options for the custody and mediation of value back to L1 are far more limited.

The option researchers reach for most is BitVM, and it is genuinely one of the better tools for the job: a way to condition Bitcoin spends from an automated, on-chain custodian through tapscript. But BitVM does not function at L0. An L0 is a fundamentally higher-level consensus program than the application logic we hoist onto Bitcoin scripting itself. It has its own consensus program. It can receive signals from a variety of sources (L1s and L2s, however diverse), and depending on the solution it may keep its own stateful ledger and mint its own assets.

An L0 may be a monolithic signer, with hardcoded code paths for how it mediates value between networks or relays oracle information across them. SUBFROST takes the harder, more general road: a single framework that works natively on Bitcoin, built on the FROST threshold-signature implementation from the Zcash Foundation.

SUBFROST as an L0

SUBFROST is one deployment of subzero-rs: a generic software framework for the federation of a Taproot custodian, with additional libraries for settlement on other cryptocurrency networks. At its core it is an off-chain multisignature wallet that supports rotation to new keys. Every holder of a share of the private key behind a given group identity runs the same WASM program. That program can make network calls, and it natively speaks frtun, the peer-to-peer network that bridges anonymized streams between the signers themselves, and out to the wider internet, where it receives upstream signals.

Crucially, SUBFROST runs a signals-based WASM program that reads from the state of Alkanes, which lives purely on L1. It runs no stateful ledger of its own. Instead it coordinates, in real time, to bring the group of signers to consensus over layers of signals, organized as a DAG, or more precisely a dependency graph.

Consensus by signals

Signals are layered in the direction of their dependencies. A SUBFROST program might first reach consensus on a single fact: the current Bitcoin block tip height. With that settled, it pulls from Alkanes to determine the set of frBTC → BTC unwraps pending for the next block. Once every signal in the graph is ascertained, each signer independently constructs the same set of PSBTs. When those sets match across the group, the system enters a signing mode, signs with the shared group key, and broadcasts the transaction that releases the outgoing BTC from the federation.

No sequencer decides the order. No single party holds the key. The agreement is the mechanism.

One framework, many groups

A subzero-rs node can manage a set of WASM (wasip2) programs and operate in more than one signing group at a time. Because the coordination logic is generic (consensus over arbitrary signals), a single node can serve several federations at once, and the SUBFROST core team runs several. The frBTC framework, the frUSD framework, and the oracle framework are all wasip2 programs, with keys distributed in exactly the manner described above.

Together they make possible assets that peg to external sources or hold reserves denominated in Bitcoin or in EVM stablecoins. They also provide price feeds drawn from real-world assets, which let us design perpetuals markets and surface genuinely useful on-chain data on Alkanes for other developers building financial structure.

The limit of Bitcoin consensus

The L0 is how we coordinate signals across every channel, and SUBFROST brings the concept to the logical extreme of what is possible on Bitcoin.

If Alkanes is the limit of what Bitcoin consensus can accomplish on L1, then SUBFROST is the limit of what Bitcoin consensus can support on L0. Together, they form the foundation for what we believe should be the final state of cryptocurrency and decentralized finance.

— flex