Home/AI Agents/MiroFish-Offline

MiroFish-Offline

nikmcfly/MiroFish-Offline

Offline multi-agent simulation and prediction engine with Neo4j and Ollama local stack

GitHub Website

GraphCanon updated Aug 14, 2026 · GitHub synced Aug 14, 2026

29views this month

2.5k stars653 forksLast push Mar 24, 2026 Python AGPL-3.0

Decision brief

MiroFish-Offline is an offline simulation and prediction engine for multi-agent systems that uses Neo4j and Ollama framework locally.

Good fit when

Use MiroFish-Offline when you require simulations in an offline environment to predict behaviors of multi-agent systems with local data storage options using Neo4j.
Employ it when your project benefits from the integration of Neo4j and Ollama, offering a robust local stack for analyzing complex relationships within simulated scenarios without internet access.

Avoid when

Avoid MiroFish-Offline if real-time agent interactions are necessary as this tool operates in an offline setup.
Do not use it when your hardware limitations fall below the recommended specifications, such as less than 32 GB RAM and less than 8 cores CPU, especially for efficient LLM inference tasks.

Requirements:: Min 16 GB RAM; Requires Docker

Observed Jul 17, 2026 · Source: enrich:decision_facts

MiroFish-Offline alternatives Trust report

Verify the decision

Maintenance and security

Full trust report

Maintenance: Slowing (142d since push); As of Aug 14, 2026
Provenance: Not a fork · Personal account; As of Aug 14, 2026
Security (OSV): 1 low (1 low); As of Jul 15, 2026

Public GitHub metadata and optional OSV scans. Signals, not a guarantee. Trust methodology.

Install

pip install MiroFish-Offline

PyPI

Similar tools

Same-category neighbours. No typed graph edges are catalogued for this tool yet.

ai-agents-for-beginners

12 Lessons to Get Started Building AI Agents

73k★

anything-llm

Self-hosted AI agent experience

66k★

ai-engineering-hub

Tutorials on LLMs, RAGs, and real-world AI agent applications

37k★

Evidence and technical details

Sourced facts, taxonomy, compatibility claims, README excerpt, and machine-readable endpoints.

Overview

An open-source Python-based offline system for simulating and predicting behaviors of multi-agent systems using a local Neo4j graph database and Ollama framework.

Capability facts

Deploy: Self-host
Source: dockerfile:Dockerfile · Aug 14, 2026
Docker: Dockerfile present
Source: dockerfile:Dockerfile · Aug 14, 2026
MCP server: No MCP server detected
Source: repo_scan · Aug 14, 2026
Languages: python, javascript
Source: github.language+package.json · Aug 14, 2026

README

Option A: Docker (easiest) `` qwen2.5:14b qwen2.5:7b`. License AGPL 3.0 — same as the original MiroFish project. See LICENSE.

For agents

This page has a .md twin and JSON over the API.

/tools/nikmcfly-mirofish-offline.md GET /api/graphcanon/tools/nikmcfly-mirofish-offline GET /api/graphcanon/graph?tool=nikmcfly-mirofish-offline