Claude Code Mcps

[](https://docs.anthropic.com/en/docs/claude-code) [](LICENSE) [](#servers)

Production-tested Model Context Protocol (MCP) servers for Claude Code CLI and Claude projects. These servers enable Claude to interact with your tools, APIs, and data sources seamlessly.

What are MCP Servers?

MCP servers extend Claude's capabilities by providing access to:

• External APIs (GitHub, WordPress, Google Search Console, GA4, LinkedIn, Gumroad)
• Databases (PostgreSQL, Redis, Notion)
• Web tools (Firecrawl web scraping, Puppeteer browser automation)
• Communication (Telegram messaging)
• File systems (local filesystem access)

Each server is configured in ~/.mcp.json and auto-loads when Claude Code starts.

Quick Install

1. Clone this repository

git clone https://github.com/arturseo-geo/claude-code-mcps.git
cd claude-code-mcps

2. Copy MCP configuration

cp mcp.json.template ~/.mcp.json

3. Configure authentication

Edit ~/.mcp.json and add API keys/credentials for the servers you want to use (see [Configuration](#configuration) below).

4. Test connection

claude code --version

Claude will load your MCP servers on startup.

Servers

| Server | Purpose | Auth | Status | |--------|---------|------|--------| | GitHub | Repos, PRs, issues, code search, actions | PAT | ✅ Prod | | WordPress | Posts, pages, media, SEO, caching | App password | ✅ Prod | | Google Search Console | Search analytics, sitemaps, URL inspection | Service account | ✅ Prod | | Google Analytics 4 | Traffic, engagement, events, conversions | Service account | ✅ Prod | | LinkedIn | People, companies, jobs, posts | Browser session | ✅ Prod | | PostgreSQL | Custom databases, queries, data operations | Connection string | ✅ Prod | | Redis | Cache, queues, sorted sets, operations | Host/port | ✅ Prod | | Notion | Pages, databases, blocks, comments | API token | ✅ Prod | | Gumroad | Products, sales, offers, email | API token | ✅ Prod | | Firecrawl | Web scraping, crawling, search, extraction | API key | ✅ Prod | | Puppeteer | Browser automation, screenshots, interaction | Local headless | ✅ Prod | | Telegram | Messages, dialogs, channels | Session token | ✅ Prod | | Filesystem | Local file read/write, directory operations | Allowed paths | ✅ Prod |

Configuration

GitHub MCP

{
  "mcpServers": {
    "github": {
      "command": "npx",
      "args": ["@modelcontextprotocol/server-github"],
      "env": {
        "GITHUB_PERSONAL_ACCESS_TOKEN": "ghp_xxxxxxxxxxxx"
      }
    }
  }
}

Get a PAT:

1. Go to https://github.com/settings/tokens
2. Click "Generate new token (classic)"
3. Scope: repo, read:user, user:email, read:org
4. Copy and paste into .mcp.json

WordPress MCP

{
  "mcpServers": {
    "wordpress": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-wordpress"],
      "env": {
        "WORDPRESS_SITES": "[{\"url\":\"https://example.com\",\"username\":\"user\",\"password\":\"pass\"}]",
        "WORDPRESS_AUTH_METHOD": "app_password"
      }
    }
  }
}

Get an app password:

1. Log in to WordPress admin
2. Go to Users → Your Profile → Application Passwords
3. Create new app password, copy username + password

Google Search Console MCP

{
  "mcpServers": {
    "gsc": {
      "command": "npx",
      "args": ["@modelcontextprotocol/server-gsc"],
      "env": {
        "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/service-account-key.json"
      }
    }
  }
}

Get service account credentials:

1. Go to Google Cloud Console
2. Create a new service account
3. Grant it "Search Console Editor" role
4. Create JSON key and save to ~/.gcp/gsc-sa.json
5. Set GOOGLE_APPLICATION_CREDENTIALS path in .mcp.json

Google Analytics 4 MCP

{
  "mcpServers": {
    "ga4": {
      "command": "npx",
      "args": ["@modelcontextprotocol/server-ga4"],
      "env": {
        "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/service-account-key.json",
        "GA4_PROPERTY_ID": "528383456"
      }
    }
  }
}

LinkedIn MCP

{
  "mcpServers": {
    "linkedin": {
      "command": "npx",
      "args": ["@modelcontextprotocol/server-linkedin"],
      "env": {
        "LINKEDIN_PROFILE": "your-profile-name"
      }
    }
  }
}

Setup:

1. Install: npm install -g @modelcontextprotocol/server-linkedin
2. Run setup: linkedin-mcp-setup (uses patchright Chromium)
3. Profile stored at ~/.linkedin-mcp/profile/

PostgreSQL MCP

{
  "mcpServers": {
    "postgres": {
      "command": "npx",
      "args": ["@modelcontextprotocol/server-postgres"],
      "env": {
        "DATABASE_URL": "postgresql://user:password@host:5432/dbname"
      }
    }
  }
}

Redis MCP

{
  "mcpServers": {
    "redis": {
      "command": "npx",
      "args": ["@modelcontextprotocol/server-redis"],
      "env": {
        "REDIS_URL": "redis://host:6379"
      }
    }
  }
}

Firecrawl MCP

{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": ["@modelcontextprotocol/server-firecrawl"],
      "env": {
        "FIRECRAWL_API_KEY": "fc-xxxxxxxxxxxx"
      }
    }
  }
}

Get API key: https://app.firecrawl.dev/api-keys

Use Cases

Content Operations

• Blog Publishing: WordPress + GitHub (version control) + Firecrawl (research) + GA4 (analytics)
• Content Pipeline: GitHub (code) + PostgreSQL (job tracking) + Redis (queue) + Telegram (notifications)

SEO & Analytics

• Search Intelligence: GSC (rankings) + GA4 (traffic) + PostgreSQL (keyword DB) + Firecrawl (competitor analysis)
• Traffic Investigations: GSC + GA4 + WordPress (content) + Notion (reporting)

Link Building & Research

• Backlink Intelligence: PostgreSQL (link data) + Firecrawl (new prospects) + LinkedIn (outreach)
• Competitor Analysis: Firecrawl (site structure) + PostgreSQL (internal DB) + Puppeteer (JS-heavy sites)

Email & Sales

• Campaign Management: Gumroad (products/sales) + Telegram (notifications) + PostgreSQL (audience)
• Lead Tracking: Notion (CRM) + GitHub (pipeline code) + Firecrawl (enrichment)

Troubleshooting

Server won't connect

1. Check ~/.mcp.json syntax: cat ~/.mcp.json | jq
2. Verify credentials are correct
3. Test individual server: npx @modelcontextprotocol/server-github --version
4. Check Claude Code logs: ~/.claude/logs/

Permission denied errors

• GitHub: Verify PAT has required scopes (see Configuration above)
• GCP: Check service account has project Editor role
• WordPress: Verify app password is correct
• PostgreSQL: Check connection string and user permissions

Slow connections

• Firecrawl: Set proxy: "enhanced" for better reliability
• PostgreSQL: Check network latency to DB server
• LinkedIn: Browser session may be stale — re-run setup

Security Best Practices

1. Never commit .mcp.json with real credentials — add to .gitignore
2. Use environment variables: export GITHUB_PERSONAL_ACCESS_TOKEN=xxxx
3. Service accounts: Use minimal roles (GSC: Search Console Editor only)
4. API keys: Rotate regularly, use different keys per environment
5. Database credentials: Use IAM/managed auth where possible
6. Terraform/IaC: Manage sensitive values via .tfvars files

Contributing

PRs welcome! To add a new server:

1. Add server entry to table (above)
2. Document configuration section
3. Include use case example
4. Update CONTRIBUTING.md
5. Test and document any quirks

See [CONTRIBUTING.md](CONTRIBUTING.md) for full guidelines.

License

MIT — see [LICENSE](LICENSE)

Built and maintained by Artur Ferreira · thegeolab.net · [MIT License](LICENSE)

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: arturseo-geo
• Source: arturseo-geo/claude-code-mcps
• License: MIT
• Homepage: https://thegeolab.net

Install and usage instructions live in the source repository linked above.