nikmcfly

nikmcfly / MiroFish-Offline

Public

Offline multi-agent simulation & prediction engine. English fork of MiroFish with Neo4j + Ollama local stack.

ai multi-agent neo4j offline ollama
121
26
100% credibility
Found Mar 16, 2026 at 116 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A fully local tool for simulating social media reactions to documents using AI agents that mimic real public discourse and opinion dynamics.

How It Works

1
🔍 Discover MiroFish

You hear about a tool that lets you test how the public might react to news or announcements on social media, all running safely on your own computer.

2
📤 Upload your document

Share a press release, report, or policy draft and describe the scenario you want to simulate, like 'how will students react?'

3
🧠 Knowledge map appears

The tool reads your document and creates a map of key people, groups, and connections, so everyone starts informed.

4
👥 Build your virtual crowd

Hundreds of unique virtual people with personalities, biases, and memories get ready to chat online just like real folks.

5
🌐 Watch the conversation unfold

Hit play and see posts, replies, arguments, and opinion shifts happen hour by hour across simulated social platforms.

6
📈 Get the smart summary

A helpful guide reviews everything, spots trends, and explains what happened with charts and insights.

💬 Talk to anyone

Chat directly with any virtual person to understand their views, backed by full memory of the whole conversation.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 116 to 121 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is MiroFish-Offline?

MiroFish-Offline is a Python-based, fully local multi-agent simulation engine that lets you upload documents like press releases or reports, then generates hundreds of AI agents to mimic public reactions on simulated Twitter and Reddit feeds. It builds a Neo4j knowledge graph from your content, creates agent personas via Ollama LLMs, runs hour-by-hour opinion dynamics, and spits out reports plus agent interviews—all offline with Docker Compose. Solves the need for private, zero-cloud prediction of social sentiment without API costs or data leaks.

Why is it gaining traction?

This English fork of the original Chinese MiroFish swaps cloud services for Ollama and Neo4j, delivering a true offline GitHub LLM setup that runs on your hardware—no keys, no internet. Developers dig the end-to-end workflow: graph build, agent sims, auto-reports, and live chats with persistent agents, plus easy Docker spins for quick tests. Stands out as an offline GitHub Copilot alternative for multi-agent prediction, beating cloud-heavy tools in privacy and speed on beefy GPUs.

Who should use this?

PR crisis teams simulating backlash to announcements, traders gauging market sentiment from news, policy wonks testing public response to drafts, or AI researchers prototyping local multi-agent swarms. Ideal for ops folks with 32GB RAM wanting offline GitHub repository analysis without vendor lock-in.

Verdict

Grab it if you need local MiroFish prediction—Docker setup is dead simple, docs solid for an early fork. At 67 stars and 1.0% credibility, it's raw but functional; expect bugs on lighter hardware, but fork it for your offline multi-agent engine.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.