Harnessmesh — Git skill for Claude Code
Persistent multi-agent collaboration fabric for agentic coding harnesses.
How to install Harnessmesh
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open domehahn/harnessmesh and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Harnessmesh does
Persistent multi-agent collaboration fabric for agentic coding harnesses. Connect Antigravity, OpenAI Codex, Claude Code and GitHub Copilot through MCP, shared channels, threads, events, evidence and durable state—without human copy/paste.
Alternatives in Git
- GitHub Review Comments — Use when a PR has unresolved review comments that need responses -- evaluates each comment, implements valid f 1.5k ★
- MCP Memory Service — GitHub Repo stars MCP server providing semantic memory and persistent storage for Claude Memory service MCP 1.5k ★
- Loki Verify — Run Loki's deterministic PR verification on the current change and summarize the evidence verdict 1.1k ★
README
HarnessMesh
**HarnessMesh is an agent-to-agent collaboration fabric for AI coding harnesses.**
- NOT an LLM proxy.
- NOT another model router.
- NOT a vendor-specific wrapper.
HarnessMesh is the **broker, control plane, and collaboration fabric** that enables diverse AI coding harnesses to communicate directly with each other in real-time through standard tools (MCP) without human copy/paste.
┌─────────────┐ ┌─────────────┐
│ Claude Code │ │ OpenAI │
│ Adapter │ │ Codex │
└──────┬──────┘ └──────┬──────┘
│ │
▼ ▼
═════════════════════════════════════
HarnessMesh
═════════════════════════════════════
▲ ▲
│ │
┌──────┴──────┐ ┌──────┴──────┐
│ Google │ │ GitHub │
│ Antigravity │ │ Copilot CLI │
└─────────────┘ └─────────────┘
HarnessMesh (Agent Collaboration Plane)
├── Collaboration Control Plane & Session Broker
├── Model Context Protocol (MCP) Server (10 Tools)
├── Peer Message Bus (`harnessmesh.peer/v1`)
├── Dynamic Adapter Registry & Lifecycle Manager
├── Capability-Based Peer Selection & Routing
├── Parallel Multi-Review & Conservative Deduplication
├── Context Projection & Secret Filtering
├── Evidence Store & Challenge Resolution
├── Loop, Depth & Budget Controller
└── Durable SQLite Session Persistence (WAL)
Switchyard (Model Routing Plane - Optional)
├── Model Selection & Provider Routing
└── Latency / Capability / Cost Routing
Switchyard is a model-routing plane. HarnessMesh is an agent-collaboration plane.
Supported First-Class Harness Adapters (v0.3.
Related Skills
Multi Agent Collab
A reusable skill for bootstrapping consistent multi-agent collaboration in any Git repository. Supports Claude
Agent Skills And Policies
Shared skills and global policies for Google Antigravity, GitHub Copilot, OpenAI Codex, and Claude Code.
The 3 Clis
Three agent-CLI harnesses (GitHub Copilot CLI, OpenAI Codex CLI, Claude Code) that take a framework pack over
Agent Intern
MCP bridge that drives four coding CLIs — Google's Antigravity (agy), OpenAI Codex, GitHub Copilot, and Cursor
JustSessions
Native macOS app to browse, search, preview, resume, and branch Claude Code, OpenAI Codex CLI, and Google Anti
Ia Dev Ebook
Um manual de campo prático sobre como desenvolver software no dia a dia trabalhando com agentes de código como
Related Agents
Developer Overview
code-container (container) creates isolated Docker environments for AI coding harnesses (Claude Code, OpenCode
Hook Entegrasyon Analisti
Claude Code ve Antigravity CLI (ve gerekirse Codex CLI / Copilot CLI) hook, izin ve headless mod davranışların
Social Ingest
Ingests external social-media research (Instagram creators / accounts, TikTok, YouTube channels, competitor ne