I build autonomous trading agents, risk systems, and direct-to-chain SDKs—tools designed to give you edge compute without sacrificing sovereignty. No fragile middle-layers, no rented infrastructure. Just keys on your machine and code you control.

The Core Principle: Deterministic code for execution, AI for judgment. Slash commands handle the fast, zero-credit operations, while LLMs handle the complex reasoning.

A full-stack, open-source Hyperliquid trading environment featuring a dedicated dashboard and a native Telegram plugin. It uses a custom conviction engine and strict slash commands for deterministic execution, completely isolating trade logic from AI hallucinations. Long-term, the architecture is designed to deeply integrate an agentic framework into the execution loop, but today it operates as a highly robust, locally hosted operational trading interface.

Built for the 2026 Hedera Apex Hackathon. A raw edge-compute execution surface for AI agents. It provides direct, own-rolled communications to Hedera mainnet, cutting out APIs that can rate-limit or rug you. The agent drives via strict CLI tools, meaning the LLM can't invent swaps or bypass read-only governance caps.

A ~400-line, zero I/O Lambda function that ships with a venue-agnostic rebalancer. No ML, no databases—just a power law, Kleiber's Law, and a heartbeat. It strictly rebalances at a 15% threshold, banking value on every swing while avoiding fee drag. Five locked constants, zero manual tuning.

A production Python SDK talking directly to SaucerSwap V2 router contracts. Built for AI agents that require deterministic, always-on execution in Hedera DeFi without risking middleware outages. This powers my own trading stack and is open-sourced for anyone building resilient on-chain agents.

_{📬 rhczs6bz4d@privaterelay.appleid.com · 🌐 portfolio · 🐙 github.com/Chris0x88 · 🤗 huggingface.co/Chris0x88}