KuljeetSinghXDA/nexter-mcp-server

Nexter MCP Server

AI-powered content management for WordPress with Nexter Blocks. This MCP (Model Context Protocol) server enables intelligent creation and modification of WordPress posts and pages built with Nexter's 96 Gutenberg blocks (duplicates merged, Pro versions prioritized).

Status: Production Ready |

Features

• 🤖 AI-Powered Content Management - Natural language commands for WordPress content
• 🎨 96 Nexter Blocks Supported - Complete schema coverage (duplicates merged, Pro versions prioritized)
• 🛡️ Safe Editing - Multi-layer validation prevents breaking existing layouts
• 📝 WordPress Revisions - Automatic revision creation for all changes
• 🔄 Rollback Support - Easy rollback via WordPress revisions
• 🎯 Smart Ambiguity Resolution - Intelligent handling of vague commands
• 📦 Docker Ready - Deploy on Dokploy with one command

Architecture

AI-Powered CLI
  ↓ MCP Protocol (HTTP/JSON-RPC 2.0)
MCP Server (nexter.aiengineops.com)
  ↓ REST API (Application Password)
WordPress (tailoredhealth.ai)
  ↓ Nexter Blocks (96 blocks)
Website Visitors

Documentation:

• - 30-minute deployment
• - Current project status
• - Detailed deployment guide
• - Complete technical spec (8000+ lines)

Quick Start

Prerequisites

• Node.js 20+
• WordPress 5.9+ with Nexter Blocks (Free + Pro)
• WordPress Application Password
• Dokploy (for production deployment)

Installation

1. Install WordPress Plugin

# Upload wordpress-plugin/nexter-mcp-api to WordPress
# Activate: Plugins → Nexter MCP API → Activate

2. Create Application Password

WordPress Admin → Users → Your Profile → Application Passwords
Application Name: "Nexter MCP Server"
Click: Add New Application Password
Copy the generated password (format: xxxx xxxx xxxx xxxx)

3. Configure Environment

cp .env.example .env
# Edit .env with your WordPress URL and credentials

4. Extract Block Schemas

IMPORTANT: Schemas must be extracted in the workspace where Nexter plugin files exist, then committed to git.

# In this workspace (has access to Nexter plugin files)
npm install
npm run extract-schemas

# This scans:
# - ../the-plus-addons-for-block-editor/classes/blocks/
# - ../the-plus-addons-for-block-editor-pro/classes/blocks/
# And generates 96 schema files in schemas/ directory

# Verify schemas created
ls schemas/*.json | wc -l
# Should show 96+ files

# Commit schemas to git
git add schemas/
git commit -m "Add extracted Nexter block schemas"

Note: Docker build will copy these pre-generated schemas. It does NOT extract them during build.

5. Run Locally (Development)

npm run dev

Server starts at http://localhost:3000

6. Deploy to Dokploy (Production)

# Push to GitHub
git init
git add .
git commit -m "Initial commit"
git remote add origin https://github.com/yourusername/nexter-mcp-server.git
git push -u origin main

# In Dokploy:
# - Create new application
# - Connect to GitHub repo
# - Set environment variables (see .env.example)
# - Configure domain: nexter.aiengineops.com
# - Deploy

Environment Variables

See for all configuration options.

Required:

• WORDPRESS_URL - Your WordPress site URL
• WP_USERNAME - WordPress admin username
• WP_APP_PASSWORD - WordPress Application Password

Optional:

• SCHEMA_REFRESH_SECRET - Secret for schema refresh endpoint
• LOG_LEVEL - Logging level (debug, info, warn, error)
• MCP_TRANSPORT - Transport mode (http for remote, stdio for local)

MCP Tools

The server exposes these tools to AI agents:

get_block_schemas

Get Nexter block schemas for specific blocks, use cases, or categories.

{
  "block_names": ["tpgb/tp-accordion", "tpgb/tp-heading"],
  "use_case": "faq-section",
  "category": "interactive"
}

create_content

Create new WordPress post or page with Nexter blocks (always as DRAFT).

{
  "post_type": "post",
  "title": "My New Post",
  "blocks": [...]
}

edit_content

Edit existing post/page with targeted operations.

{
  "post_id": 456,
  "operations": [
    {
      "operation": "modify_block",
      "block_id": "test001",
      "new_attrs": {"title": "Updated Title"}
    }
  ]
}

analyze_content

Analyze post/page to identify blocks and structure.

{
  "post_id": 456,
  "include_schemas": true
}

search_content

Search for posts or pages.

{
  "query": "acne blog",
  "post_type": "post",
  "limit": 10
}

validate_content

Validate block structure before saving.

{
  "blocks": [...],
  "strict": false
}

Project Structure

nexter-mcp-server/
├── src/                      # TypeScript source code
│   ├── index.ts             # Server entry point
│   ├── tools/               # MCP tool implementations
│   ├── services/            # Core services (WordPress client, schema loader)
│   ├── utils/               # Utilities (logger, validators)
│   └── resources/           # MCP resources
├── schemas/                  # Block schemas (generated)
├── scripts/                  # Utility scripts
│   └── extract-schemas.js   # Schema extraction
├── wordpress-plugin/         # WordPress plugin
│   └── nexter-mcp-api/      # Upload this to WordPress
├── docs/                     # Documentation
├── Dockerfile               # Docker build configuration
├── docker-compose.yml       # Dokploy deployment
└── package.json             # Dependencies

Development

# Install dependencies
npm install

# Run in development mode
npm run dev

# Build for production
npm run build

# Run tests
npm test

# Lint code
npm run lint

# Extract schemas (after Nexter updates)
npm run extract-schemas

Testing

# Test health check
curl http://localhost:3000/health

# Test with WordPress
curl -X POST http://localhost:3000/test-connection

# View logs
tail -f logs/combined.log

Deployment

Dokploy Configuration

1. Create Application in Dokploy
2. Connect GitHub repository
3. Set Environment Variables (from .env.example)
4. Configure Domain: nexter.aiengineops.com
5. Enable SSL: Let's Encrypt via Traefik
6. Deploy

Dokploy automatically:

• Builds Docker image
• Creates isolated network
• Configures Traefik routing
• Provisions SSL certificate
• Starts container with health checks

Security

• WordPress Application Passwords (revocable)
• All credentials in environment variables
• No hardcoded secrets
• Input validation on all endpoints
• Rate limiting enabled
• HTTPS enforced via Traefik

Monitoring

• Health check endpoint: /health
• Structured logging with Winston
• Request/response logging
• Error tracking
• Performance metrics (optional)

Maintenance

Update Schemas

When Nexter Blocks is updated, extract schemas locally in this workspace and push to GitHub:

# 1. Update Nexter plugin files in workspace
# Copy latest versions to:
#   - ../the-plus-addons-for-block-editor/
#   - ../the-plus-addons-for-block-editor-pro/

# 2. Re-extract schemas in this workspace
cd nexter-mcp-server
npm run extract-schemas

# 3. Verify schemas updated
ls schemas/*.json | wc -l  # Should be 96+
git diff schemas/  # Review changes

# 4. Commit and push to GitHub
git add schemas/
git commit -m "Update schemas for Nexter vX.X.X"
git push origin main

# 5. Dokploy auto-redeploys on push
# Or manually: Dokploy → nexter-mcp-server → Redeploy

Why this workflow?

• ✅ Workspace has access to Nexter plugin source files
• ✅ Schemas are version-controlled and reviewable
• ✅ Docker build is fast (just copies pre-generated schemas)
• ✅ No need for Nexter plugins in Docker image

View Logs

# In Dokploy
docker logs nexter-mcp-server --tail=100 -f

# Local
tail -f logs/combined.log

Troubleshooting

Can't Connect to WordPress

# Test WordPress REST API
curl https://your-site.com/wp-json/

# Test Application Password
curl -u "username:app-password" https://your-site.com/wp-json/wp/v2/users/me

# Check plugin is activated
# WordPress Admin → Plugins → Nexter MCP API should be active

Schema Extraction Fails

# Verify Nexter plugins exist in workspace
ls ../the-plus-addons-for-block-editor/classes/blocks/
ls ../the-plus-addons-for-block-editor-pro/classes/blocks/

# Run extraction with verbose output
node scripts/extract-schemas.js

Docker Build Fails

# Build locally to debug
docker build -t nexter-mcp-test .

# Check logs
docker logs nexter-mcp-server

Documentation

• - 30-minute deployment guide
• - Current project status and features
• - Complete deployment instructions
• - Full technical specification (8000+ lines)
• - Plugin documentation

License

MIT

Author

Built for Tailored Health AI project - WordPress content management automation.