Chrysalis v2 Overview

Developer documentation for Chrysalis v2, a cross-chain USDC settlement and protocol-execution system using route planning, risk policy, and bridge adapters.

Where Chrysalis v1 answers "how do I stake from a given network," v2 answers a broader question: given a source chain, a destination chain, an amount, and a target protocol, plan and execute the entire path as one atomic pipeline - bridge, then act - and hand back a receipted, auditable record of what happened and why. Instead of a user manually bridging and then separately calling a protocol, they submit a single intent - move USDC from Arc to Base, then supply it to Aave - and the platform's agents evaluate it, pick a route, execute it on-chain, and mint a receipt.

How it works

1. An intent is submitted - source chain, destination chain, amount, target protocol, and an optimization goal (balanced, lowest_cost, fastest, or safest).
2. Agents evaluate and quote it. A RiskPolicyAgent checks allowlists, slippage bounds, and transfer limits; a FeeQuoteAgent prices every eligible bridge route and scores them against the stated goal.
3. A RoutePlannerAgent picks the route and builds a step-by-step pipeline: verify deposit -> bridge/mint -> call the protocol adapter -> narrate the receipt.
4. The route executes on-chain through ArcIntentRouterUpgradeable and a protocol adapter on the destination chain, and Chrysalis mints a receipt NFT on Arc recording the full flow.

Five product pillars

• Arc as the settlement hub - native USDC gas, EURC, Circle Gateway, CCTP V2, BridgeKit, Circle Paymaster, and EIP-3009/x402 nanopayment flows all anchor to Arc.
• Pluggable protocol execution - every DeFi protocol sits behind an isolated adapter contract or program; adding one never requires a core router upgrade or state migration.
• Upgradeable control plane - the EVM router (ArcIntentRouterUpgradeable) is UUPS-upgradeable, and can register, pause, or swap adapters independently of contract logic.
• AI agent automation - policy-aware agents plan routes, verify compliance, quote fees pre-flight, autonomously rebalance treasury liquidity, and generate human-readable audit narrations.
• Quote-first execution - gas, bridge fees, protocol fees, slippage, and paymaster sponsorship are all computed and shown, with alternatives and rejection reasons, before a signature is requested.