harness-sync

Synchronize AI coding tool configurations across Claude Code, Cursor, and GitHub Copilot.

9|2|Updated Mar 7, 2026
One-click install
npx skills add https://github.com/harnessprotocol/harness-kit --skill harness-sync-harnessprotocol
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: harness-sync
Source: https://github.com/harnessprotocol/harness-kit/tree/main/plugins/harness-share/skills/harness-sync
Command: npx skills add https://github.com/harnessprotocol/harness-kit --skill harness-sync-harnessprotocol

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents your Claude Code, Cursor, and GitHub Copilot setups from drifting out of sync so you don’t lose updates, miss MCP servers, or end up with conflicting instruction blocks across platforms.

Core Features & Use Cases

  • Inventories across platforms: Detects installed skills, harness instruction blocks, and MCP server configurations on each supported tool.
  • Divergence reporting with conflict awareness: Identifies what’s missing or different and surfaces conflicts side-by-side without silent overwrites.
  • Bidirectional syncing with guardrails: Supports push (reusing harness-compile logic) or pull (merging native changes back into harness.yaml) while requiring explicit user confirmation.

Quick Start

Run /harness-sync to scan your project, show divergence, and then follow the prompts to sync in the direction you choose.

Frequently Asked Questions about harness-sync

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I keep AI coding tool configs aligned across Claude Code, Cursor, and GitHub Copilot?

You can keep AI coding tool configs aligned by running a config sync that inventories skills, harness instruction blocks, and MCP server configurations across Claude Code, Cursor, and GitHub Copilot to detect and resolve divergence.

How does MCP server configuration divergence reporting work across multiple AI coding tools?

Divergence reporting works by scanning installed MCP server configurations, skills, and instruction blocks across your supported platforms, surfacing missing or conflicting settings side-by-side without silent overwrites.

Can I pull native configuration changes back into a central harness.yaml file?

Yes, you can pull native changes back into harness.yaml by merging tool configurations bidirectionally, which requires explicit user confirmation before modifying the file to ensure safe workflows.

Does config sync silently overwrite conflicting harness instruction blocks during a push?

No, config sync does not silently overwrite conflicts. It detects unresolved issues, skips them, and reuses harness-compile logic for push writes while requiring confirmation for pull operations.

What is the best way to sync MCP server configs and skills across different AI platforms?

The best way to sync MCP server configs is to run a bidirectional synchronization workflow that inventories configurations across platforms, generates a divergence report, and applies guarded push or pull actions.