Bitcoin's limited programmability and transaction throughput have historically prevented native Bitcoin from participating in decentralized finance (DeFi) applications. Existing solutions depend on honest-majority thresholds, or centralized custodial entities that introduce significant trust requirements. This paper introduces Bitcoin Smart Accounts (BSA), a novel protocol that enables native Bitcoin to access DeFi through trust-minimized infrastructure while maintaining self-custody of funds. BSA achieves this through a combination of emulated Bitcoin covenants using Partially Signed Bitcoin Transactions (PSBTs) and Taproot scripts, a Trusted Execution Environment (TEE)-based arbitration system, and destination chain smart contracts that enable DeFi platforms to accept self-custodial Bitcoin as collateral without necessitating protocol-level modifications. The setup leverages liquidity secured by the Lombard Security Consortium which provides a twofold advantage: for a DeFi protocol, liquidators rely on fungible assets with deep liquidity to quickly exit positions, while for a depositor, the general trust assumptions of honest majority (m-of-n) are reduced to existential honesty (1-of-k). We present the complete protocol design, including the Bitcoin architecture, the TEE-based arbitration mechanism, and the Smart Account Registry for protocol management. We provide a security analysis that demonstrates the correctness, safety, and availability properties under our trust model. Our design enables native Bitcoin to serve as collateral in lending markets and other DeFi protocols without requiring users to relinquish custody of funds.
Redactable Blockchain in the Permissionless Setting
比特币是一种不可变的非许可区块链系统，已被许多严重依赖其不可变性的不同应用程序广泛用作公共公告板。然而，比特币的不变性并非没有其公平份额的缺点。国际刑警组织揭露了比特币区块链中存在有害和潜在非法的文件，图像和链接，此后对目前存在于比特币区块链中的数据类型进行了几次定性和定量分析。尽管人们对区块链有很多关注，但令人惊讶的是，以前为无权限设置中的数据编辑提出的解决方案远不可行，并且需要额外的信任假设。因此，有害数据的问题仍然对国际刑警组织等执法机构构成巨大挑战（Tziakouris，IEEE S& P '18）。我们提出了第一个高效的可编辑区块链，用于无需许可的设置，可以很容易地集成到比特币中，并且不依赖于繁重的加密工具或信任假设。我们的协议使用基于共识的投票，并由一个策略参数化，该策略规定了编辑的要求和约束;如果编辑收集了足够的投票，则在链上执行操作。作为一个额外的功能，我们的协议为编辑链提供了公共可验证性和问责制。此外，我们提供了正式的安全定义和证明，表明我们的协议是安全的，对未达成共识的编辑。此外，我们展示了我们的方法的可行性与一个概念验证的实现，显示只有一个很小的开销，在我们的协议的链验证相比，一个不变的。
