Building at the intersection of onchain finance and AI agent infrastructure.

Dynamic risk-fee Uniswap v4 Hook for new-asset, meme, and low-cap pools on X Layer. LP fees scale with swap impact — Normal 0.30%, Warning 1.00%, Protected 3.00% — and the pool's risk state is verifiable onchain, so traders and LPs price the same risk the contract does.

Ships with a demo router, a live frontend, and a Foundry test suite covering the fee override, risk decay, and hard-threshold blocking paths.

Solidity Foundry Uniswap v4 X Layer — live demo

The quality layer for agents that ship code. Point it at a public repo and get evidence-backed findings (every claim carries a path:line:reason pointer) and rule-based reproducible scoring — plus a fix plan per finding: ordered steps, tests to add, acceptance criteria, and an agentInstructions block you can hand straight to Codex or Claude Code.

Re-audit and see what actually changed: score movement attributed rule by rule, findings split into resolved / new / still-open. The loop is audit → fix plan → fix → re-audit → compare, over HTTP or MCP.

Static analysis only — it never executes the audited repository's code.

TypeScript Fastify MCP x402

• AI agent infrastructure — memory, skills, harness design, MCP tooling
• DeFi primitives — Uniswap v4 Hooks, onchain risk pricing, X Layer
• Developer tooling — the kind that actually ships

中文：做链上金融与 AI Agent 基础设施的交叉地带。上面两个项目，一个在 X Layer 上做 Uniswap v4 动态风险费率 Hook；另一个是给 Agent 用的仓库质量层——审计出问题、给出可执行的修复计划、改完再审计做前后对比。