sqlanon-mcp

cerby0n/sqlanon-mcp

3.2

If you are the rightful owner of sqlanon-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The SQLite Anonymization MCP Server is a specialized server designed to anonymize SQLite databases, providing tools for data anonymization with detailed reporting.

Tools
4
Resources
0
Prompts
0

SQLite Anonymization MCP Server

A Model Context Protocol (MCP) server specialized in SQLite database anonymization. This server provides tools to preview and execute data anonymization with comprehensive reporting.

Features

  • Query Preview: See exactly what SQL queries will be executed before making changes
  • Multiple Anonymization Strategies: 12 different strategies including masking, hashing, categorization, and more
  • Backup Support: Automatically backup databases before anonymization
  • Comprehensive Reporting: Detailed reports of all queries executed and rows affected
  • Docker Support: Run in isolated containers for security
  • Database Inspection: Explore database schema before anonymization
  • Advanced Features: Pattern masking, date extraction, numeric categorization, and column deletion

Anonymization Strategies

StrategyDescriptionExampleParameters
hashReplace with random hex stringabc123... → f4e3d2c1...-
randomizeReplace with random stringJohn Doe → a7b3f9e2-
nullSet value to NULLJohn Doe → NULL-
fixedSet to a fixed valueJohn Doe → REDACTEDvalue
maskReplace with masking patternJohn Doe → ***pattern
emailGenerate random emailuser@example.com → a7b3@example.com-
phoneGenerate random phone555-1234 → 555-7382-
keep_first_char_maskKeep first char + mask restMike → M*********maskChar, maskLength
random_digitsRandom N-digit number12345 → 9384756201digitCount
extract_yearExtract year from date01.01.1999 → 1999-
categorizeBucket numeric values75000 → mediumcategories[]
delete_columnRemove column entirelyColumn deleted-

Installation

Local Installation

npm install
npm run build
npm start

Docker Installation

# Build the Docker image
docker build -t sqlite-anonymization-mcp .

# Or use docker-compose
docker-compose up -d

Usage with Claude Desktop

Add to your Claude Desktop configuration (claude_desktop_config.json):

{
  "mcpServers": {
    "sqlite-anonymization": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-v",
        "/path/to/your/databases:/databases",
        "-v",
        "/path/to/backups:/backups",
        "sqlite-anonymization-mcp"
      ]
    }
  }
}

For local (non-Docker) usage:

{
  "mcpServers": {
    "sqlite-anonymization": {
      "command": "node",
      "args": ["/path/to/sqlanon/build/index.js"]
    }
  }
}

Available Tools

1. inspect_database

Inspect database schema to understand tables and columns.

Input:

{
  "dbPath": "/databases/mydata.db"
}

Output:

{
  "dbPath": "/databases/mydata.db",
  "tables": ["users", "orders"],
  "schema": {
    "users": [
      {"cid": 0, "name": "id", "type": "INTEGER", "notnull": 0, "dflt_value": null, "pk": 1},
      {"cid": 1, "name": "email", "type": "TEXT", "notnull": 1, "dflt_value": null, "pk": 0},
      {"cid": 2, "name": "name", "type": "TEXT", "notnull": 1, "dflt_value": null, "pk": 0}
    ]
  }
}

2. preview_anonymization

Preview anonymization queries before execution.

Input:

{
  "dbPath": "/databases/mydata.db",
  "rules": [
    {
      "table": "users",
      "column": "email",
      "strategy": "email"
    },
    {
      "table": "users",
      "column": "name",
      "strategy": "hash"
    }
  ]
}

Output:

{
  "message": "Preview of anonymization queries (NOT executed)",
  "dbPath": "/databases/mydata.db",
  "queries": [
    {
      "index": 1,
      "table": "users",
      "column": "email",
      "strategy": "email",
      "query": "UPDATE users SET email = lower(hex(randomblob(8))) || '@example.com' WHERE email IS NOT NULL"
    },
    {
      "index": 2,
      "table": "users",
      "column": "name",
      "strategy": "hash",
      "query": "UPDATE users SET name = lower(hex(randomblob(16))) WHERE name IS NOT NULL"
    }
  ],
  "totalQueries": 2,
  "note": "Use execute_anonymization to apply these changes"
}

3. execute_anonymization

Execute anonymization on the database.

Input:

{
  "dbPath": "/databases/mydata.db",
  "backupPath": "/backups/mydata_backup.db",
  "rules": [
    {
      "table": "users",
      "column": "email",
      "strategy": "email"
    },
    {
      "table": "users",
      "column": "phone",
      "strategy": "phone"
    },
    {
      "table": "users",
      "column": "ssn",
      "strategy": "null"
    }
  ]
}

Output:

{
  "message": "Anonymization completed successfully",
  "report": {
    "timestamp": "2025-10-06T15:30:00.000Z",
    "dbPath": "/databases/mydata.db",
    "queries": [
      {
        "query": "UPDATE users SET email = lower(hex(randomblob(8))) || '@example.com' WHERE email IS NOT NULL",
        "affectedRows": 150,
        "table": "users",
        "column": "email",
        "strategy": "email"
      },
      {
        "query": "UPDATE users SET phone = '555-' || substr(cast(abs(random()) as text), 1, 4) WHERE phone IS NOT NULL",
        "affectedRows": 150,
        "table": "users",
        "column": "phone",
        "strategy": "phone"
      },
      {
        "query": "UPDATE users SET ssn = NULL WHERE ssn IS NOT NULL",
        "affectedRows": 150,
        "table": "users",
        "column": "ssn",
        "strategy": "null"
      }
    ],
    "totalQueries": 3,
    "totalRowsAffected": 450
  },
  "backupCreated": true,
  "backupPath": "/backups/mydata_backup.db"
}

4. get_anonymization_report

Retrieve the report for a previously anonymized database.

Input:

{
  "dbPath": "/databases/mydata.db"
}

Output: Returns the full anonymization report from the last execution.

Example Workflow with Claude

  1. Inspect the database:

    "Use the inspect_database tool on /databases/users.db"

  2. Preview anonymization:

    "Preview anonymization for the users table:
- email column with email strategy
- name column with hash strategy
- ssn column with null strategy"

  3. Review the queries that Claude presents

  4. Execute anonymization:

    "Execute the anonymization with backup to /backups/users_backup.db"

  5. Get the report:

    "Show me the anonymization report for /databases/users.db"

Security Considerations

  • Always create backups before anonymization
  • Run in Docker for isolation
  • Test on a copy of your database first
  • Review preview queries carefully before execution
  • Store backups securely
  • Consider using read-only database mounts when inspecting

Project Structure

sqlanon/
├── src/
│   └── index.ts          # Main MCP server implementation
├── build/                # Compiled JavaScript (generated)
├── databases/            # Database files (mounted volume)
├── backups/              # Backup storage (mounted volume)
├── package.json          # Node.js dependencies
├── tsconfig.json         # TypeScript configuration
├── Dockerfile            # Docker container definition
├── docker-compose.yml    # Docker Compose configuration
└── README.md            # This file

Development

# Install dependencies
npm install

# Build
npm run build

# Run locally
npm start

# Build Docker image
docker build -t sqlite-anonymization-mcp .

Troubleshooting

Database not found

  • Ensure the database path is absolute
  • When using Docker, ensure volumes are mounted correctly
  • Check file permissions

Permission denied

  • Ensure the database file is writable
  • Check Docker volume permissions
  • Run with appropriate user permissions

Queries not working as expected

  • Use preview_anonymization to see exact queries
  • Use inspect_database to verify table and column names
  • Check SQLite version compatibility

License

MIT

Contributing

Contributions welcome! Please open an issue or submit a pull request.