skillcp
Syncs skills and MCP server configurations to GitHub Copilot and VS Code, using their global and project skill directories and MCP config files.
Syncs skills and MCP server configurations to OpenAI Codex using its skills directories and codex configuration files.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@skillcpadd the github MCP server and sync it to all my tools"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Skillcp
Yes. Skills and MCP servers are already portable — every major harness just hides them in a different folder and JSON shape. Skillcp is the missing library: one copy of each skill and MCP server, published to every harness you use.
It is a CLI, an MCP server, an Agent Skill, and a local web UI, so Cursor, Claude Code, Copilot, Codex, Gemini CLI, Windsurf, OpenCode, Cline, Claude Desktop, and Pi can all drive it — or you can manage the library in a browser.
What it does
Keeps a canonical library in
~/.skillcpSkills as standard
SKILL.mdpackages (~/.skillcp/skills/<name>/)MCP servers as one
mcp.json(~/.skillcp/mcp.json)
Imports whatever you already configured in Cursor, Claude, Copilot, and the rest
Syncs that library back out:
Skills are symlinked (copied if the OS blocks links) into each product's skills directory
MCP servers are merged into each product's native config format without clobbering unrelated settings
After sync, the same GitHub MCP server and the same code-review skill show up in every harness.
Related MCP server: Lab Registry Server
Install
npm install -g skillcp
skillcp --versionOr run without installing:
npx skillcpFrom a checkout:
npm install
npm run build
node dist/cli.js statusFirst run:
skillcp init --import --install --syncThat creates the library, vacuums up existing skills/MCP configs, registers Skillcp itself as a skill + MCP server, and writes everything back to the harnesses it detects.
Everyday use
# Browser GUI for the library
skillcp ui
# See what is installed where
skillcp status
skillcp doctor
# Add a skill from a folder or GitHub repo
skillcp skill add ./my-skill
skillcp skill add anthropics/skills --path pdf
# Add an MCP server once
skillcp mcp add github --command npx --args -y,@modelcontextprotocol/server-github --env GITHUB_TOKEN=$GITHUB_TOKEN
skillcp mcp add docs --url https://example.com/mcp --type http
# Publish the library to every detected harness (user-global)
skillcp sync
# Also write project files (.cursor/skills, .mcp.json, .vscode/mcp.json, …)
skillcp sync --project
# Only Cursor + Claude Code
skillcp sync --to cursor --to claudeUse it from any harness (the plugin)
skillcp install adds:
the
skillcpAgent Skill (so the model knows when to use it)the
skillcpMCP server (skillcp serveover stdio)
Any MCP-capable harness can then call tools like skillcp_add_skill, skillcp_add_mcp, skillcp_import, skillcp_sync, and skillcp_open_ui instead of editing five config files by hand.
This repo is also a native plugin for the IDEs:
Cursor / VS Code / Copilot / Windsurf
code --install-extension ./extensions/vscode
# or, in Cursor: Install from VSIX / install the folder as an extensionThe extension registers the Skillcp MCP server, contributes the skill, and adds commands: Open UI, Sync, Import, Status.
Claude Code
This repository is a Claude Code plugin (/.claude-plugin/plugin.json). Enable it with:
claude plugin install .
# or
claude --plugin-dir .If the MCP server is not connected, agents can still run the CLI. For a browser GUI, run skillcp ui. The bundled skill at skills/skillcp/SKILL.md documents both paths.
Web UI
skillcp uiStarts a small local web server (default http://127.0.0.1:8787) and opens it in your browser. From there you can create, edit, and remove skills, add MCP servers, import from installed harnesses, and sync the library back out. It binds to localhost only.
skillcp ui --port 9000 --no-openSupported harnesses
Harness | Skills (global / project) | MCP config (global / project) |
Cursor |
|
|
Claude Code |
|
|
GitHub Copilot / VS Code |
|
|
Windsurf |
|
|
OpenAI Codex |
|
|
Gemini CLI |
|
|
OpenCode |
|
|
Cline | — |
|
Claude Desktop | — |
|
Generic |
| — |
Pi |
|
|
MCP conversion covers the real format drift: mcpServers vs servers vs OpenCode mcp vs Codex TOML mcp_servers, plus Gemini's httpUrl vs url.
Pi has no native MCP; Skillcp still writes mcp.json in Pi's agent dir so pi-mcp-adapter (and similar extensions) pick the same servers up.
Several hosts scan more than their own skill folder (Cursor also reads Claude, Codex, and .agents; Copilot and Gemini read .agents; Pi reads .agents). Sync writes each skill to the fewest folders that still reach every product, and removes leftover Skillcp links that would show up twice. MCP configs are per-product and are not folded this way.
skillcp harnesses prints the resolved paths on your machine.
How a library looks
~/.skillcp/
config.json # symlink vs copy
mcp.json # canonical MCP servers
manifest.json # names Skillcp manages
skills/
code-review/SKILL.md
skillcp/SKILL.md
backups/ # copies of harness files before each MCP writeSync is additive. Extra MCP servers a harness already had are left alone. Servers Skillcp used to manage that left the library are dropped from harness files on the next sync (and immediately on skill rm / mcp rm). --force replaces skill folders that are not Skillcp symlinks. --keep on rm leaves harness copies in place.
Project vs global
Default sync is user-global, so every repo you open gets the same personal skills and MCP servers.
--project writes into the current working directory for team-shared setup. Commit those files if the whole team should get them.
Secrets
Skillcp does not rewrite or encrypt env values. Prefer environment-variable references ($GITHUB_TOKEN) in mcp.json instead of pasting tokens. Backups under ~/.skillcp/backups will contain whatever was already in the harness file.
CLI
skillcp init [--import] [--install] [--sync]
skillcp status [--project]
skillcp doctor
skillcp import [--to cursor] [--all] [--project] [--overwrite]
skillcp sync [--to claude] [--all] [--project] [--dry-run] [--force] [--prune]
skillcp install [--no-sync]
skillcp serve
skillcp ui [--port 8787] [--host 127.0.0.1] [--no-open]
skillcp skill list|add|rm|show
skillcp skill rm <name> [--keep] [--project]
skillcp mcp list|add|rm|show
skillcp mcp rm <name> [--keep] [--project]
skillcp harnessesWhat we deliberately do not do
Skillcp stays a local library plus adapters. It does not:
Store secrets in a keychain (copy env values as written; prefer
$ENV_NAME)Cover every editor (Amp, Roo, Continue, Zed, Goose, Crush, …). New hosts are a small adapter if you need them.
skill rm and mcp rm unsync detected harnesses by default. Pass --keep to leave those copies alone (skills that were symlinks are copied first so they keep working).
skillcp with no arguments prints a short status and the two commands that matter: ui and sync.
Development
npm install
npm test
npm run build
node dist/cli.js statusNode 20+ is required.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseNot gradedqualityAmaintenanceEnables AI agents to discover, install, and manage SKILL.md skills from a Git-backed registry via MCP tools for search, install, and list operations.51MIT
- FlicenseAqualityAmaintenanceEnables MCP clients to discover and fetch skills, agents, commands, and hooks from the Gen-e2 Lab Registry, and check local inventory compliance against the registry.15
- AlicenseNot gradedqualityBmaintenanceDiscovers and manages portable agent capabilities (skills and MCP servers) from configurable collections, providing search, inspection, and local installation via CLI and MCP tools.1MIT
- AlicenseNot gradedqualityBmaintenanceEnables managing local repositories, CLIs, MCP servers, and skills via MCP tools, with inventory tracking, guarded updates, and Docker-sandboxed trials.MIT
Related MCP Connectors
A registry of 5,900+ peer-authored skills any MCP agent can search and load on demand.
Search your team's shared AI-skill library, get install commands, and save skills from your agent.
Manage SRG+ hubs, channels, content, assets, users, and workspaces from any MCP-aware AI agent.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/lum1n/skillcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server