cloak.business MCP Server

Zero-setup MCP integration for AI tools. Connect to cloak.business's PII anonymization directly from Cursor, Claude Desktop, or any MCP-compatible client.

Supports 48 languages, regional entity groups, presets, and custom recognizers.

Quick Start

Option 1: Streamable HTTP (Cursor, Web Clients)

No Node.js required - just add the URL to your configuration.

{
  "mcpServers": {
    "cloak-business": {
      "url": "https://cloak.business/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Option 2: stdio Transport (Claude Desktop)

See CLAUDE_DESKTOP.md for full setup guide.

{
  "mcpServers": {
    "cloak-business": {
      "command": "npx",
      "args": ["-y", "cloak-business-mcp-server"],
      "env": {
        "cloak_business_API_KEY": "your-api-key-here"
      }
    }
  }
}

Getting Your API Key

1. Sign up at cloak.business
2. Go to Dashboard > API Keys
3. Create a new API key
4. Copy and use in your MCP configuration

Available Tools

All tool names are prefixed with cloak_business_ for namespace clarity.

cloak_business_analyze_text

Analyze text for PII without modifying it.

Parameters: | Parameter | Type | Description | |-----------|------|-------------| | text | string | Required. The text to analyze | | output | string | "full" | "summary" | "counts" (default: "full") | | language | string | Language code - 48 supported (default: "en") | | entities | string[] | Specific entity types to detect | | entity_groups | string[] | Entity groups: UNIVERSAL, FINANCIAL, DACH, etc. | | score_threshold | number | Confidence threshold 0.0-1.0 (default: 0.5) | | preset | string | "default" | "personal" | "private" | | ad_hoc_recognizers | object[] | Custom regex-based recognizers |

Example:

{
  "text": "Contact John Doe at [email protected], SSN 123-45-6789",
  "language": "en",
  "entity_groups": ["UNIVERSAL", "NORTH_AMERICA"],
  "score_threshold": 0.7,
  "output": "counts"
}

Response:

{
  "entity_counts": { "PERSON": 1, "EMAIL_ADDRESS": 1, "US_SSN": 1 },
  "entities_found": 3,
  "tokens_charged": 8,
  "score_threshold": 0.7
}

cloak_business_anonymize_text

Cloakize text by replacing PII with placeholders or reversible tokens.

Parameters: | Parameter | Type | Description | |-----------|------|-------------| | text | string | Required. The text to anonymize | | mode | string | "redact" (permanent) | "tokenize" (reversible) | | output | string | "full" | "metadata" | "minimal" | | language | string | Language code - 48 supported (default: "en") | | entities | string[] | Specific entity types to detect | | entity_groups | string[] | Entity groups to detect | | score_threshold | number | Confidence threshold 0.0-1.0 | | preset | string | "default" | "personal" | "private" | | persistence | string | "session" (24h) | "persistent" (30d) | | sessionName | string | Optional name for the session | | operators | object | Custom operators per entity type |

Operators Example:

{
  "text": "John Doe, email: [email protected], card: 4111-1111-1111-1111",
  "mode": "redact",
  "operators": {
    "PERSON": { "type": "replace", "new_value": "[REDACTED NAME]" },
    "EMAIL_ADDRESS": { "type": "hash", "hash_type": "SHA256" },
    "CREDIT_CARD": { "type": "mask", "chars_to_mask": 12, "from_end": false }
  }
}

Response:

{
  "anonymized_text": "[REDACTED NAME], email: a1b2c3d4e5f6..., card: ************1111",
  "entities_found": 3,
  "tokens_charged": 12
}

cloak_business_detokenize_text

Restore original PII values from tokenized text.

Parameters:

• text (required): The tokenized text
• sessionId (required): Session ID from anonymization
• partial: Allow partial restoration (default: false)

cloak_business_analyze_image

Analyze an image for PII using OCR. Returns detected entities with bounding box positions.

Parameters: | Parameter | Type | Description | |-----------|------|-------------| | image | string | Required. Base64-encoded image data | | image_format | string | Required. "png" | "jpeg" | "bmp" | "tiff" | | language | string | OCR language hint, 2-letter ISO code (default: "en") | | entities | string[] | Specific entity types to detect | | score_threshold | number | Confidence threshold 0.0-1.0 (default: 0.5) |

Example:

{
  "image": "<base64-encoded-png>",
  "image_format": "png",
  "language": "en"
}

Response:

{
  "entities": [
    { "entity_type": "PERSON", "score": 0.85, "left": 24, "top": 31, "width": 159, "height": 35 },
    { "entity_type": "EMAIL_ADDRESS", "score": 1.0, "left": 25, "top": 81, "width": 284, "height": 35 }
  ],
  "entities_found": 2,
  "tokens_charged": 3,
  "processing_time_ms": 260
}

cloak_business_redact_image

Redact PII from an image using OCR. Returns a base64-encoded redacted image.

Parameters: | Parameter | Type | Description | |-----------|------|-------------| | image | string | Required. Base64-encoded image data | | image_format | string | Required. "png" | "jpeg" | "bmp" | "tiff" | | language | string | OCR language hint, 2-letter ISO code (default: "en") | | entities | string[] | Specific entity types to redact | | score_threshold | number | Confidence threshold 0.0-1.0 (default: 0.5) | | fill_color | string | "black" | "white" | "red" | "green" | "blue" | "gray" (default: "black") |

Response:

{
  "redacted_image": "<base64-encoded-png>",
  "redacted_image_format": "png",
  "image_size_bytes": 4313,
  "tokens_charged": 3,
  "processing_time_ms": 279
}

cloak_business_batch_analyze

Analyze multiple texts for PII in a single request. Supports 1-100 texts with parallel processing.

Parameters: | Parameter | Type | Description | |-----------|------|-------------| | items | array | Required. Array of texts to analyze (1-100 items) | | items[].text | string | Required. The text to analyze | | items[].id | string | Optional identifier for this item | | items[].language | string | Language code for this item (default: "en") | | items[].entities | string[] | Specific entity types to detect | | output | string | "full" | "summary" | "counts" (default: "full") |

Example:

{
  "items": [
    { "id": "msg-1", "text": "John Smith at [email protected]", "language": "en" },
    { "id": "msg-2", "text": "Hans Müller, IBAN DE89370400440532013000", "language": "de" }
  ],
  "output": "counts"
}

Response:

{
  "results": [
    { "id": "msg-1", "success": true, "entity_counts": { "PERSON": 1, "EMAIL_ADDRESS": 1 }, "metadata": { "language": "en", "entities_found": 2, "tokens_used": 2 } },
    { "id": "msg-2", "success": true, "entity_counts": { "PERSON": 1, "IBAN_CODE": 1 }, "metadata": { "language": "de", "entities_found": 2, "tokens_used": 2 } }
  ],
  "batch_metadata": {
    "total_items": 2,
    "successful": 2,
    "failed": 0,
    "total_entities_found": 4,
    "total_tokens_used": 4,
    "processing_time_ms": 320
  }
}

cloak_business_get_balance

Check your current token balance. Free operation.

cloak_business_list_sessions

List your active tokenization sessions. Free operation.

cloak_business_delete_session

Delete a tokenization session. Free operation.

Supported Languages (48)

spaCy Languages (25)

| Code | Language | Code | Language | Code | Language | |------|----------|------|----------|------|----------| | en | English | de | German | es | Spanish | | fr | French | it | Italian | pt | Portuguese | | nl | Dutch | pl | Polish | ru | Russian | | ja | Japanese | zh | Chinese | ko | Korean | | ro | Romanian | el | Greek | hr | Croatian | | sl | Slovenian | mk | Macedonian | sv | Swedish | | da | Danish | nb | Norwegian | fi | Finnish | | uk | Ukrainian | lt | Lithuanian | ca | Catalan | | tr | Turkish | | | | |

Stanza NER Languages (7)

| Code | Language | Code | Language | |------|----------|------|----------| | bg | Bulgarian | hu | Hungarian | | he | Hebrew | vi | Vietnamese | | af | Afrikaans | hy | Armenian | | eu | Basque | | |

Transformer Languages (16)

| Code | Language | Code | Language | Code | Language | |------|----------|------|----------|------|----------| | ar | Arabic | hi | Hindi | cs | Czech | | sk | Slovak | id | Indonesian | th | Thai | | fa | Persian | sr | Serbian | lv | Latvian | | et | Estonian | ms | Malay | bn | Bengali | | ur | Urdu | sw | Swahili | tl | Tagalog | | is | Icelandic | | | | |

Entity Groups

Use entity_groups parameter to detect regional entities easily:

| Group | Entities Included | |-------|-------------------| | UNIVERSAL | PERSON, EMAIL_ADDRESS, PHONE_NUMBER, LOCATION, DATE_TIME, URL, IP_ADDRESS, etc. | | FINANCIAL | CREDIT_CARD, IBAN_CODE, SWIFT_CODE, CRYPTO | | DACH | DE_TAX_ID, AT_SSN, CH_AHV, and more for Germany, Austria, Switzerland | | FRANCE | FR_NIR, FR_CNI, BE_NATIONAL_NUMBER, LU_NATIONAL_ID | | SPAIN_LATAM | ES_NIF, ES_NIE, MX_CURP, AR_DNI, CL_RUT, etc. | | ITALY | IT_FISCAL_CODE, IT_VAT, IT_DRIVER_LICENSE, IT_HEALTH_CARD | | PORTUGAL_BRAZIL | PT_NIF, PT_CC, BR_CPF, BR_CNPJ, BR_RG | | NORDIC | DK_CPR, FI_HETU, SE_PERSONNUMMER, NO_FODSELSNUMMER, IS_KENNITALA | | POLAND | PL_PESEL, PL_NIP, PL_REGON, PL_PASSPORT | | UK_IRELAND | UK_NHS, UK_NINO, IE_PPS, UK_PASSPORT | | NORTH_AMERICA | US_SSN, US_PASSPORT, CA_SIN, CA_HEALTH_NUMBER | | ASIA_PACIFIC | JP_MY_NUMBER, CN_RESIDENT_ID, SG_NRIC, AU_TFN, IN_AADHAAR | | MIDDLE_EAST | AE_EMIRATES_ID, SA_NATIONAL_ID, IL_ID_NUMBER | | HEALTHCARE | MEDICAL_LICENSE, NRP, HEALTH_INSURANCE_ID |

Custom Recognizers (Ad-hoc)

Define custom patterns for domain-specific entities:

{
  "text": "Employee ID: EMP-12345, Badge: B-9876",
  "ad_hoc_recognizers": [
    {
      "entity_type": "EMPLOYEE_ID",
      "patterns": [{ "regex": "EMP-\\d{5}", "score": 0.9 }],
      "context": ["employee", "staff", "worker"]
    },
    {
      "entity_type": "BADGE_NUMBER",
      "patterns": [{ "regex": "B-\\d{4}", "score": 0.85 }]
    }
  ]
}

Operator Types

| Operator | Description | Parameters | |----------|-------------|------------| | replace | Replace with custom value | new_value | | redact | Replace with <ENTITY_TYPE> | - | | hash | One-way hash | hash_type: SHA256, SHA512 | | encrypt | Reversible encryption | key: 16/24/32 char key | | mask | Partial masking | chars_to_mask, masking_char, from_end | | keep | Keep original value | - |

Architecture

This MCP server supports two transports:

• Streamable HTTP: For web clients, Cursor IDE - processing on cloak.business servers
• stdio: For Claude Desktop - local process, API calls to cloak.business

Your data is:

• Encrypted in transit (TLS 1.3)
• Processed in memory only (not stored)
• Tokenization mappings encrypted at rest (AES-256-GCM)
• GDPR compliant (EU data residency)

Version History

• v2.6.1 - Security hardening: error sanitization, request ID propagation
• v2.6.0 - Batch analyze: process 1-100 texts in a single request
• v2.5.0 - Retry logic with exponential backoff for transient failures
• v2.4.0 - Structured logging, enhanced health checks, error codes
• v2.3.0 - Image anonymization: analyze and redact PII in images via OCR
• v2.2.1 - Timing-safe API key comparison to prevent timing attacks
• v2.2.0 - 48 languages, presets, entity groups, advanced features
• v2.1.0 - stdio transport for Claude Desktop
• v2.0.x - Streamable HTTP transport, enhanced tools
• v1.0.0 - Initial release with stdio transport

Support

• Website: cloak.business
• Documentation: cloak.business/docs
• Claude Desktop Setup: CLAUDE_DESKTOP.md
• Support: [email protected]

License

MIT License