Building autonomous AI agents, smart contract security tools, Solana programs and various other stuff. Some of this work is built by my autonomous AI agent, Aurora, who runs 24/7 on a dedicated server contributing to open-source projects, occasionally competing in hackathons, and conducting security research.
Trustless agent-to-agent payment protocol on Solana. AI agents can transact, verify, and settle — no human intermediary.
Everything you need to make an AI autonomous. One file. Production-tested across 700+ sessions.
An autonomous AI agent — runs 24/7 on a dedicated server doing security research, open-source contributions, and agent commerce. Self-governing with economic decision-making, adversarial review, and mechanical enforcement.
| Project | Description |
|---|---|
| multi-strategy-vault | ERC-4626 tokenized vault with multi-strategy yield allocation |
| aurora-treasury | Multi-chain wallet management for autonomous AI agents |
| stablepay-protocol | Multi-sig USDC vault with compliance controls on Solana |
| medicaid-fraud-detector | Fraud signal detection engine for Medicaid provider data |
| solana-tx-decoder | Decode Solana transactions into human-readable format |
Active contributor to open-source projects including TypeORM, ProjectDiscovery (tlsx, httpx), and various Solana ecosystem tools. Aurora handles PR contributions, bug fixes, and security audits autonomously.
Aurora writes a blog about autonomous AI, security research, and building in the open: marchantdev.github.io/blog
X: @MarchantDev