skills.rest
madebywild
madebywildOFFICIAL
Explore all skills by this author →
WebsiteGitHub

harness-mcp

Merge MCP server JSON files and generate provider-specific artifacts.

12|Updated Feb 6, 2026
Search Tags:#mcp#workspace#merge#harness#config-management#server-config#provider-output
One-click install
npx skills add https://github.com/madebywild/agent-harness --skill harness-mcp-madebywild
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: harness-mcp
Source: https://github.com/madebywild/agent-harness/tree/main/.claude/skills/harness-mcp
Command: npx skills add https://github.com/madebywild/agent-harness --skill harness-mcp-madebywild

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Consolidates scattered MCP server definitions into a single, merge-ready source of truth.

Core Features & Use Cases

  • MCP definitions live as JSON files under .harness/src/mcp. Multiple files merge at harness apply time to form a unified configuration.
  • Collision safety: if the same server ID appears in more than one file, the plan is rejected to prevent overwrites.
  • Provider output mapping: merged MCP data is rendered into the provider-specific formats used by Claude Code, OpenAI Codex, and GitHub Copilot.

Quick Start

Run harness add mcp to scaffold a new MCP config, then edit the file under .harness/src/mcp and execute harness apply to generate provider outputs.

Frequently Asked Questions about harness-mcp

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

FAQPage Schema
How do I manage multiple MCP server configs for different AI coding assistants?▼

MCP server definitions are maintained as JSON files under .harness/src/mcp and merged at apply time into a unified configuration, which is then rendered into provider-specific formats for Claude Code, OpenAI Codex, and GitHub Copilot.

What happens if I have duplicate MCP server IDs across different config files?▼

When duplicate server IDs appear across multiple MCP config files, the apply plan is automatically rejected to prevent silent overwrites and configuration collisions. This collision safety mechanism ensures your merged MCP configuration remains consistent and predictable.

Does this approach support generating configs for Claude Code, OpenAI Codex, and GitHub Copilot?▼

Yes, merged MCP data is rendered into provider-specific formats for Claude Code, OpenAI Codex, and GitHub Copilot through defined mappings. This allows you to maintain one unified source while outputting compatible artifacts for each AI coding assistant.

How do I set up and apply centralized MCP server configurations?▼

Run harness add mcp to scaffold a new config file, edit the JSON under .harness/src/mcp, then execute harness apply to merge definitions and generate provider outputs. This workflow ensures your MCP entity structure is validated and rendered correctly.

Why do I need to centralize MCP server definitions instead of keeping them separate?▼

Centralizing MCP server definitions prevents configuration drift across multiple AI coding assistants by providing a single, merge-ready source of truth. It enforces consistent entity structure and automates provider output generation, eliminating manual synchronization across separate config files.

Related Skills

Explore More Software Engineering Skills→
thisrohanguptathisrohangupta
1

create-environment

Generate Harness environment YAML configurations and push them via MCP.

Community
Intermediate
madebywildmadebywild
12

harness-settings

Centralize provider-specific harness settings across Claude Code, OpenAI Codex CLI, and GitHub Copilot.

Official
Intermediate
thisrohanguptathisrohangupta
1

create-infrastructure

Automate Harness infrastructure definition YAML and deployment targets via MCP v2.

Community
Intermediate
SHAJAR5110SHAJAR5110

chatkit-backend

Deploy a FastAPI ChatKit backend with multi-provider AI integration.

Community
Intermediate