MiroFish: A Swarm Intelligence Engine for Predicting Anything

MiroFish is a next-generation AI prediction engine built on multi-agent technology. By extracting seed information from the real world, it automatically constructs high-fidelity parallel digital worlds. Within these worlds, thousands of agents with independent personalities, memories, and reasoning capabilities interact with each other, helping you rehearse the future before it happens.

Core Features

• Graph Construction — Real-world seed extraction, memory injection, and GraphRAG building.
• Environment Setup — Entity-relationship extraction, persona generation, and parameter injection.
• Social Evolution Simulation — Dual-platform parallel simulation with dynamically updated temporal memory for accurate forecasting of future trends.
• Report Generation — A deep-interaction ReportAgent environment that produces comprehensive analytical reports.
• Deep Interaction — Conversational access to any agent, allowing fine-grained exploration of the digital world.

Application Scenarios

• Macro Decision-Making — A rehearsal laboratory for decision makers. Test policies and PR strategies in a zero-risk digital sandbox.
• Micro Creative Exploration — A creative sandbox for individuals. Rehearse novel endings, explore infinite "what-ifs," and brainstorm wild ideas.

Workflow

1. Upload Seed — Provide initial real-world context or data. 2. Describe Need — Specify the scenario, question, or hypothesis to simulate. 3. Get Report — Receive a detailed forecast generated by the multi-agent simulation.

Tech Stack

• Python Backend
• Vue Frontend
• OASIS Engine (multi-agent simulation core)

Open Source

MiroFish is released as an open-source project.

• Repository: github.com/666ghj/MiroFish