skills.rest
reason-machines
reason-machinesOFFICIAL
Explore all skills by this author →
Website𝕏@reasonmachinesGitHub

mcp-servers-reference

Catalogs 50 vetted MCP servers with verified install commands and security guidance.

7|2|Updated May 16, 2026
Search Tags:#security#AI agents#claude code#MCP#model context protocol#server configuration#install commands
One-click install
npx skills add https://github.com/reason-machines/mcp-skills --skill mcp-servers-reference-reason-machines
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: mcp-servers-reference
Source: https://github.com/reason-machines/mcp-skills/tree/main/skills/mcp-servers-reference
Command: npx skills add https://github.com/reason-machines/mcp-skills --skill mcp-servers-reference-reason-machines

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? With over 20,000 MCP servers in public registries, most abandoned or insecure, developers struggle to identify trustworthy servers and configure them correctly across Claude Code, Codex, and Gemini CLI. ## Core Features & Use Cases - Curated Server Catalog: 50 vetted MCP servers organized across 9 categories including core development, databases, infrastructure, fintech, web3, trading, and creator tools. - Verified Install Commands: Copy-paste installation patterns for Claude Code, Codex, and Gemini CLI with required environment variables and known gotchas. - Security Best Practices: Token scoping guidance, sandbox credential rules, and a pre-install security checklist to prevent production accidents. - Use Case: A full-stack developer asks which MCP servers to install for a new project and receives a lean 5-server setup (GitHub, Postgres, Playwright, Sentry, Vercel) with exact commands and read-only token recommendations. ## Quick Start Ask the agent to recommend and install the best MCP servers for your workflow, such as web development or trading.

Frequently Asked Questions about mcp-servers-reference

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

FAQPage Schema
How do I add an MCP server to Claude Code?▼

Use the claude mcp add command with the server name and launch command, for example claude mcp add github npx -y @modelcontextprotocol/server-github. Hosted servers use the --transport http flag with the server URL instead.

What MCP servers should I install for web development?▼

The recommended web development setup is GitHub, PostgreSQL, Playwright, Sentry, and Vercel. Keep the total to five or fewer servers, since tool bloat from too many servers degrades agent reasoning quality.

How many MCP servers can I use at once?▼

Limit yourself to 5-7 servers maximum, ideally 3-5 matching your actual workflow. Each server exposes dozens of tools, and more than about 10 active servers overwhelms the agent and slows responses.

Does MCP server configuration work with Codex and Gemini CLI?▼

Yes. Codex uses codex mcp add or the ~/.codex/config.toml file, while Gemini CLI uses the mcpServers object in ~/.gemini/settings.json. Both accept the same command, args, and env structure as Claude.

Why does my MCP server fail with authentication errors?▼

Authentication failures usually mean an expired token, insufficient scopes, or an environment variable that was not loaded. Verify the token with curl, confirm its scopes match the server requirements, and restart the agent to reload env vars.

Is it safe to connect MCP servers to production databases?▼

No. Database servers like PostgreSQL have full write access, so point them only at dev or staging environments. Use read-only tokens, test credentials for fintech and trading servers, and never enable writes against production from an agent loop.

Related Skills

Explore More Software Engineering Skills→
mqzkimmqzkim

Play Console Setup

Guide Google Play Console app setup and service account creation for API access.

Community
Intermediate
global-osglobal-os

vercel-agent

Automate PR analysis, incident investigation via logs and metrics, and SDK installation for Vercel users.

Official
Advanced
thefear078thefear078
15

agent-design

Design and review LLM agents covering tool design, loops, memory, guardrails, and multi-agent orchestration.

Community
Advanced
atrawogatrawog

chat

Access Ollama REST API endpoints from Python for model management.

Community
Intermediate