Private cross-chain bridge with unlinkable settlement

What this enables

• →Move between chains without linking wallet A to wallet B. Ethereum wallet A deposits, the assets route through the Starknet privacy layer and private sub-account settlement, Solana wallet B receives - and the A↔B relationship is not publicly observable on either side. Every existing bridge writes both legs to transparent ledgers, so anyone can join deposit to withdrawal. Here the link is broken by construction.
• →Starknet as the privacy routing layer between transparent chains. Ethereum, Solana, Base, and Arbitrum are all publicly traceable; the moment value crosses between them on a normal bridge, the graph is complete. STRK20 and Private Sub-Accounts sit in the middle as an anonymity set - assets enter from any chain, mix, and exit to any chain with no carried-through link. The bigger the pool, the stronger every crossing.
• →A fresh destination wallet with no history. Withdraw to a brand-new address on the target chain that has no onchain connection to the funding source. Useful for post-CEX-withdrawal privacy, portfolio restructuring, treasury operations, and separating a public identity from a working wallet - without trusting a custodian to forget.
• →Compliance-compatible, not a mixer. Unlike sanctioned mixers, this keeps a disclosure path: Viewing Keys register the crossing, an authorized trace is possible under legal process, and source-of-funds can be proven selectively. Confidentiality from public surveillance - not evasion.
• →One deposit, any chain out. The user picks a source, an amount, and a destination chain; routing, settlement, and privacy are abstracted. No manual bridging, no wrapped-asset juggling, no understanding of Starknet required.

What you build

A Bridge helper contract that orchestrates deposit → shield → route → unshield across chains. Assets enter from Ethereum, Solana, Base, or Arbitrum via Chain Abstraction and land in the Privacy Pool as Encrypted Notes; settlement routes through Private Sub-Accounts so the deposit leg and the withdrawal leg share no observable link. The user signs once on the source chain with their existing wallet; the destination is a fresh address on the target chain with no onchain path back to the funding source. Timeout-and-reclaim covers the failure path. A Viewing-Key registration layer preserves selective disclosure and authorized tracing, so the bridge is compliance-compatible rather than a mixer - the anonymity set grows with every crossing, and each crossing strengthens the ones around it.

Why this isn't Houdini Swap

Houdini Swap has processed $1.5B+ in private cross-chain volume and proves the demand - people will pay to break the link between source and destination. But Houdini routes through intermediary assets and an operator that sees both ends; the privacy is custodial and trust-based. Here the routing layer is a verifiable Cairo contract and a cryptographic anonymity set, not a company that could log, be subpoenaed, or be shut down. Same job - unlinkable cross-chain movement - done without a trusted middle.

Why this isn't Near Intents

Near Intents settles cross-chain intents by having solvers fill a stated outcome, and it's efficient - but the settlement legs are public and the source-to-destination relationship remains observable on the transparent chains it touches. Privacy isn't in the model; it optimizes routing, not unlinkability. This bridge makes non-linkability the primitive: the crossing passes through STRK20 as a privacy layer, so the A↔B relationship never becomes public in the first place.

Why this isn't Jumper

Jumper (LI.FI) is best-in-class transparent bridging - it aggregates routes across dozens of bridges and chains for the cheapest, fastest path, and every hop is fully traceable. It's built to move assets, not to hide the connection between where they came from and where they went. Here chain abstraction serves privacy rather than just price: the point isn't the optimal route, it's that no one can reconstruct the route or link the endpoints.