gonostudy305

Omniagent OS — AI skill for Claude Code

AI community

A unified local control plane for AI coding agents (Claude Code, Codex, Antigravity) - cross-agent skills SSOT, drift auditing, and memory coordination.

How to install Omniagent OS

This entry records only its repository, not the path inside it, so there is no exact command to give. Open gonostudy305/omniagent-os and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Omniagent OS does

A unified local control plane for AI coding agents (Claude Code, Codex, Antigravity) - cross-agent skills SSOT, drift auditing, and memory coordination.

Alternatives in AI

  • Mex — Persistent project memory for AI coding agents 672 ★
  • Claudexor — Multi-harness control plane for Claude Code, Codex, Cursor, and OpenCode: quota-aware rotation across multiple 425 ★
  • Prismor — Self-hosted runtime control plane for AI agents 334 ★

README

OmniAgent-OS / Local Agent Skills SSOT

**A Unified Single-Source-of-Truth (SSOT) & Cross-Agent Management Layer for Claude Code, OpenAI Codex, and Antigravity (Gemini).**

[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE) [![Platform](https://img.shields.io/badge/Platform-Windows%20%7C%20macOS%20%7C%20Linux-lightgrey)](#) [![Agents Supported](https://img.shields.io/badge/Agents-Claude%20Code%20%7C%20Codex%20%7C%20Antigravity-orange)](#)


📌 The Problem: The "Siloed Agent" Nightmare

Modern developers increasingly leverage multiple AI coding engines on the same local workstation:

  • Claude Code for architecture analysis and planning.
  • OpenAI Codex CLI for rapid implementation, unit tests, and refactoring.
  • Google Antigravity (Gemini) for multimodal tasks, generative UI, and complex workflows.

However, each agent installs its own fragmented set of skills and isolated memory silos:

  • .claude/skills
  • .codex/skills & .agents/skills
  • .gemini/config/skills

🔍 Real-World Audit Findings (60,000+ Files Sample)

During our empirical audit of a multi-agent developer workstation, we observed:

  1. The Impact of Embedded Dependencies (node_modules):
    Copying skill directories across agents ballooned local storage to 60,000+ files (~1.1 GB). In our audit sample, over 85% of these files were nested node_modules inside 7–9 complex skills. These are reproducible dependencies that should be installed on demand rather than duplicated across agent configurations. The actual instructional markdown and source code accounted for ~3,200 files (~55–75 MB).
  2. Artificial "Content Drift" (Hardcoded Paths):
    Out of 38 skills flagged as differing across agent directories, 31 skills (81.6%) were logically identical; their only difference was hardcoded directory paths (e.g., .claude/skills/... vs .agents/skills/... vs .Codex/secrets/...).
  3. True Version Drift:
    Only 3 skills in