Are there specific coding rules for new agents?

Yes, agents must be kept under 800 lines of code, use the ModelFactory for LLM access, and store all outputs in the designated src/data/ directory.

What exchanges are supported by this skill?

The system currently supports Hyperliquid (perpetuals), Solana/BirdEye (spot), and Extended Exchange (X10), with modular functions for each.

What is the RBI agent?

The Research-Based Inference (RBI) agent uses DeepSeek-R1 to extract strategy logic from external sources like YouTube or PDFs and generates executable backtesting code.

Which AI models can I use with these agents?

The ModelFactory supports Claude, GPT-4, DeepSeek, Groq, Gemini, and local models via Ollama, allowing you to choose the best model for each specific task.

Does this skill help with risk management?

Yes, it includes a dedicated Risk Agent that evaluates conditions before trades and supports global circuit breakers like maximum daily loss and gain limits.

Moon Dev AI Trading Agents

Name: Moon Dev AI Trading Agents
Author: Microck

byMicrock

•

数据科学与机器学习

Orchestrates over 48 specialized AI agents for autonomous cryptocurrency trading, market analysis, and backtesting across multiple exchanges.

关于

This skill provides expert integration with Moon Dev's experimental AI trading system, a modular framework featuring over 48 autonomous agents. It enables developers to deploy, modify, and manage trading bots on Hyperliquid, Solana (BirdEye), and Extended Exchange (X10). By utilizing a unified ModelFactory, users can swap between LLMs like Claude, GPT-4, and DeepSeek to power sentiment analysis, whale tracking, and automated risk management. It is particularly useful for traders looking to bridge the gap between AI reasoning and real-time market execution while maintaining strict architectural standards.

主要功能

81 GitHub stars
RBI Agent for automated backtesting from YouTube URLs, PDFs, or text descriptions
Unified ModelFactory for seamless LLM abstraction (Claude, OpenAI, DeepSeek, Groq)
Multi-exchange support for Hyperliquid, Solana (BirdEye), and Extended Exchange
Modular architecture with standalone executables and strict file size limits for maintainability
Orchestration of 48+ specialized agents for trading, sentiment, and risk management

使用场景

Converting complex trading strategies into Python backtests using AI-driven inference
Building and deploying autonomous crypto trading bots with multi-LLM decision logic
Automating market research and sentiment analysis via specialized research agents

关于

主要功能

81 GitHub stars
RBI Agent for automated backtesting from YouTube URLs, PDFs, or text descriptions
Unified ModelFactory for seamless LLM abstraction (Claude, OpenAI, DeepSeek, Groq)
Multi-exchange support for Hyperliquid, Solana (BirdEye), and Extended Exchange
Modular architecture with standalone executables and strict file size limits for maintainability
Orchestration of 48+ specialized agents for trading, sentiment, and risk management

使用场景

Converting complex trading strategies into Python backtests using AI-driven inference
Building and deploying autonomous crypto trading bots with multi-LLM decision logic
Automating market research and sentiment analysis via specialized research agents