Skip to content

Mnehmos/Mnehmos

Repository files navigation

Mnehmos

🌐 Live Site: mnehmos.github.io/Mnehmos

MCP Infrastructure Developer · Multi-Agent Systems · AI-Powered Games

Building tools that let AI agents interact with complex stateful systems—whether that's a computer desktop, a distributed workflow, or a D&D campaign.

GitHub Stars Multi-Agent Framework MCP Tools


🔧 The Mnehmos Ecosystem

MCP Servers

Project Tools Description
mnehmos.ooda.mcp 62 Full computer control—filesystem, shell, screen, keyboard, mouse, clipboard
mnehmos.synch.mcp 19 Persistent memory bank, context handoffs, locks for agent coordination
mnehmos.trace.mcp 11 Schema validation between MCP producers and consumers
mnehmos.index-foundry.mcp 40+ Deterministic RAG pipeline → deployed chatbots in minutes
mnehmos.arxiv.mcp 4 Academic paper search with PDF extraction and caching
mnehmos.rpg.mcp 145 Full D&D 5e backend—combat, spells, quests, 1100+ creature presets
mnehmos.chatrpg.game 30+ Lightweight D&D 5e server for web deployment

Frameworks & Research

Project Description
mnehmos.multi-agent.framework 528⭐ · Production-ready multi-agent coordination (SPARC + Boomerang + TDD)
mnehmos.prompts.research 182 techniques from 20+ papers · Interactive taxonomy, prompt builder

Shipped Products

Product Platform Description
mnehmos.quest-keeper.game Desktop (Tauri) AI Dungeon Master + 3D battlemap + 7-layer context (~5K tokens)
mnehmos.long-hall.game Web Procedural roguelike with ironman saves & cloud leaderboards
mnehmos.portal-chess.game Web Multiplayer chess with portal teleportation mechanics
mnehmos.invoicer.app Web Local-storage invoice generator with PDF export

📊 By The Numbers

300+  MCP tools across 7 servers
528   GitHub stars on Multi-Agent Framework
182   Prompt engineering techniques documented
145   D&D 5e tools in RPG Engine
800+  Passing tests in game engine
1100+ Creature presets for instant encounters
7     Context layers in Quest Keeper AI

🏗️ Architecture Philosophy

Tools don't think, they act.

Every MCP server I build follows these principles:

  • Deterministic: Same inputs → same outputs
  • Idempotent: Re-running produces identical results
  • Auditable: Every operation produces manifests and logs
  • Composable: Tools can be run independently or chained

The LLM proposes. The engine validates. State comes from the database, not hallucination.


🔗 Quick Links

MCP Servers

Live Demos

Documentation


📫 Contact

  • GitHub: @Mnehmos
  • Location: Arizona, USA
  • Focus: MCP servers, multi-agent systems, AI-powered applications

Building infrastructure for AI agents that interact with the real world.

About

GitHub Profile README

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published