Arti — Tor's official Rust client — compiled to WebAssembly. Anonymized RPC from wallets, frontends, and dApps, running entirely in the browser.
Ship Arti-backed RPC routing as a general transport for wallets, SDKs (ethers.js, viem.js), and light clients — wallet-agnostic, not tied to any single client. Kohaku is the Q2 reference integration.
Merge async / time compatibility fixes back into Arti. WASM-readiness plumbing lands in Arti; the JS wrapper stays external.
Browser-native WebRTC transport replacing WebSocket for Arti↔relay. In building it we saw a larger opportunity: a general-purpose transport for reaching anonymity and p2p networks from the edge (browsers, wallets). That broader effort is now the Abstract Access Layer, with WebRTC as the core.
Internal audit of the WASM-compiled Arti client using an LLM-based review swarm; ahead of external audit (includes dedicated audit for RustCrypto lib).
External audit of the WASM-compiled Arti client before broad wallet-SDK adoption.
Land Arti-backed routing in mainstream wallet SDKs (ethers.js, viem.js) and additional light clients (Helios, etc.) for network-level-private RPC.
Survey messenger apps/protocols and propose Arti-backed anonymization integrations; pursue one reference integration.
Support edge (wallets, light clients, ..) to P2P transaction broadcasting and IP-leakage prevention, leveraging WebRTC work in TorJS plus (a) .onion-exposing infra providers (dRPC, Flashbots, …) and (b) direct ethp2p peers after that abstraction is implemented.
Use our own PIR to privately retrieve the Tor directory on bootstrap, removing a plaintext metadata leak at the start of every session.
| KPI | Target | Status |
|---|---|---|
| Integrations | 10+ wallets/sdk's/light clients | Ongoing |
| Bootstrap time | ~3 seconds | Achieved (was ~3 minutes) |
| Upstream Arti PRs | 5+ landed | Under review with Tor Project |
| Audit | External audit complete | Internal Audit 2Q26 (LLM swarm) |