@laoyingkhq: Guys, recently I saw a real beast. An undergrad in quantitative trading spent ten days building an AI cluster simulation framework, posted it on GitHub and it blew up — 13,000+ stars and $4 million in funding. That thing is MiroFish, a multi-agent simulation tool for financial prediction, public opinion testing...

X AI KOLs Timeline Tools
open-source multi-agent simulation ai-framework github quant fintech

Summary

Introduces the open-source multi-agent simulation framework MiroFish, developed by undergraduate Guo Hanjiang in ten days while still in school. It gained 13,000+ GitHub stars and $4 million in funding, and can be used for financial prediction, public opinion testing, etc. The post also promotes an automated trading bot on Polymarket.

Brothers, recently I saw a real beast. An undergrad student working in quant spent ten days building an AI cluster simulation framework, posted it on GitHub and it blew up — over 13,000 stars and $4 million in funding. That thing is MiroFish, a multi-agent simulation and reasoning tool. It can handle financial forecasting, public opinion testing, and more. In simple terms, it builds a virtual world and drops a few thousand AI agents inside. Each one has its own "brain" — they chat, take sides, and play games. You throw in an event — whether it's a policy, news, or half a novel — and the system simulates how people react, how sides argue, and how it all ends, letting you see the outcome in advance. The founder, Guo Hanjiang, was still an undergrad at the end of 2025. He used large models to help write code, and after open-sourcing it, it became wildly popular. Shanda Group directly gave 30 million yuan. Later he dropped out to start a business and continued with his team. The main use cases fall into three categories: Finance: feed in earnings reports and news, simulate market panic and price movements, and assist decision-making. PR: run announcements through the simulation first to check for negative risks, optimize, then release. Entertainment: toss in an unfinished novel and the AI writes the ending for you. Deployment is simple too — one-click Docker, and it can connect to mainstream large model APIs. You can try simulating Musk posting Dogecoin 2.0 and see how institutions, social media, and retail investors react. You can even export a video and experience market fluctuations with zero risk. Some guys are already making money with it. Someone used MiroFish to backtest historical market moves, then bet on SPX500 on Polymarket, netting over $120,000. Trader page: https://polymarket.com/zh/@CamelUp?via=YINGGE888… If you don't want to do your own research, just use PolyCop to copy trades — directly replicate what the pros do. Click this bot: https://t.me/PolyCop_BOT?start=ref_YINGGE888…, enter the wallet 0x17559efac103ac7f361be37ec0b93888d4c55aac, and you're in. Open-source code: https://github.com/666ghj/MiroFish
Original Article
View Cached Full Text

Cached at: 08/06/26, 04:33 AM

Guys, I recently came across a serious player.

An undergrad in quantitative finance spent ten days building his own AI cluster simulation framework. Posted it to GitHub and it blew up — over 13,000 stars and $4 million in funding secured.

That thing is MiroFish, a multi-agent simulation tool. It can handle financial forecasting, public opinion testing, and more.

Simply put, it builds a virtual world and drops thousands of AI agents in it. Each one has its own mind — they chat, pick sides, and play games.

You throw in an event — policy, news, or a half-written novel — and the system simulates how the crowd reacts, how parties argue, and how it all ends, letting you see the outcome in advance.

The founder, Guo Hanjiang, was still an undergrad at the end of 2025. He used LLMs to help write code. After going open source, it went wildly viral, and Shanda Group directly invested 30 million yuan.

Later he dropped out to start a company and continued building with his team.

Main use cases fall into three categories:

Finance: feed in earnings reports and news to simulate market panic and price movements, helping with decisions. PR: run announcements through simulation first to check for negative risk, then optimize before releasing. Entertainment: drop in an unfinished novel and the AI writes the ending for you.

Deployment is simple too — one-click Docker. It works with mainstream LLM APIs.

You can try simulating Elon Musk tweeting Dogecoin 2.0, see how institutions, self-media, and retail investors react, and even export video — zero risk, feel the market swings.

Some folks are already making money with it. After using MiroFish to backtest historical market movements, one person bet on SPX500 on Polymarket and netted over $120,000.

Trader page: https://polymarket.com/zh/@CamelUp?via=YINGGE888…

If you don’t want to do your own research, just use PolyCop to copy trades and directly mirror the pros’ moves.

Click this bot https://t.me/PolyCop_BOT?start=ref_YINGGE888…, enter the wallet 0x17559efac103ac7f361be37ec0b93888d4c55aac, and you’re in.

Open source code: https://github.com/666ghj/MiroFish

Polymarket @camelup

Source: https://polymarket.com/zh/@camelup Adventure One QSS Inc. ©2026·Privacy (https://polymarket.com/zh/privacy)·Terms of Use (https://polymarket.com/tos)·Market Integrity (https://integrity.polymarket.com/)·Help Center (https://help.polymarket.com/)·Docs (https://docs.polymarket.com/)

Polymarket operates globally through independent legal entities. Polymarket US (https://polymarket.us/) is operated by QCX LLC d/b/a Polymarket US, which is a CFTC-regulated Designated Contract Market. This international platform is not regulated by the CFTC and operates independently. Trading involves significant risk of loss. Please refer to our Terms of Service (https://polymarket.com/tos) and Privacy Policy (https://polymarket.com/privacy). This translation is for reference only. If there are any differences between the English text and this translation, the English version prevails.

Crypto老鹰 (@laoyingkhq): Many players in the crypto space are discussing the real state of Polymarket: ordinary participants are still stuck manually refreshing the board and staring at price movements, while automated trading scripts of all kinds have long been fully open-sourced on GitHub.

Here are 8 free automated programs across different tracks. Each open-source project comes with complete English deployment docs and detailed feature descriptions:

  1. CloddsBot

Similar Articles

@cevenif: Guys, a library literally exploded today—nearly 90k stars, called TradingAgents, a multi-agent trading framework. In plain English: a group of AIs work together to help you trade crypto and stocks. ① Some AIs specialize in monitoring market data ② Some AIs are responsible for strategy decisions ③ Some AIs specialize…

X AI KOLs Timeline

TradingAgents is a multi-agent trading framework that uses multiple AIs working together to automate crypto and stock trading, supporting real-time data ingestion, automatic strategy generation, and continuous optimization. It has gained nearly 90k stars on GitHub.

@NXR_NIROX: > A Chinese girl > no quant degree, no Bloomberg terminal, no fund background > just a laptop, opened Claude Code > spent a weekend building a bot for five instruments > mean reversion on indices, breakout on Bitcoin, trend following on commodities > ATR for position sizing, 1% hard stop loss, plus…

X AI KOLs Timeline

A Chinese girl with no quant background used Claude Code over a weekend to create a trading bot covering five instruments, employing mean reversion, breakout, and trend strategies, achieving significant returns, demonstrating the potential of AI tools in automated trading.

@Bitcoin188: Holy crap! The top minds on the internet have open-sourced their brains! These 11 GitHub repos can save you three years of detours, making freeloaders fly! Brothers, bookmark them now, study them slowly, don't waste time searching manually! PilotDeck (OpenBMB) Deploy an AI Agent that works on its own in minutes, a true open-source agent framework!

X AI KOLs Timeline

Recommend 11 high-quality open-source projects on GitHub covering AI agent frameworks, AI programming, memory systems, research automation, and quantitative investment tools, designed to help developers get started quickly and boost efficiency.

@CycleDecoded: Unbelievable, guys—GitHub just quietly unleashed a monster project that's set to disrupt institutional quant traders' livelihoods. The project is called Alpha-Dojo/DojoAgents, with a brutally hardcore positioning: 'Full-market Personal Investment AI Copilot.' This isn't some grandstanding large model that just talks in circles; it's the real deal…

X AI KOLs Timeline

Alpha-Dojo/DojoAgents is a full-market personal investment AI copilot framework that supports automated analysis and strategy backtesting for stocks, forex, and cryptocurrencies. It is fully open-source and aims to give retail investors access to institutional-grade quantitative tools.

@waveking1314: Holy cow, there are a ton of ridiculously free projects on GitHub. Many of them can already replace the software you're paying for monthly. 1. TradingAgents AI multi-agent quantitative trading framework https://github.com/TauricResearch/…

X AI KOLs Timeline

Recommends 10 high-quality GitHub open-source projects covering quantitative trading, AI chat, video generation, financial terminal, short video creation, email assistant, voice cloning, OSINT analysis, AI skill library, and API integration platform, many of which are free alternatives to paid software.