Skip to main content
Glama
VibeLeading

mcp-privacy-shield

by VibeLeading

mcp-privacy-shield

The Privacy Shield — the Sanitization Gate, Tiered Sovereignty, and Privacy HUD. Keeps identity out of the AI's hands, keeps the Vault on-premise, and monitors a live Compliance Score.

Implements the Privacy Shield server from the appendix of Vibe Leading The AI.

What it does

  • Sanitization Gate — replaces PII with tokens like [EMAIL_01] (or category masks like [EMAIL]). The AI understands the logic but never sees the identity.

  • Tiered Sovereignty — classifies data into public / internal / restricted / vault zones. Restricted PII stays in the Vault on-premise; you bring the AI to the data, not the reverse.

  • Privacy HUD — monitors a Compliance Score (0-100). Snap the connection if sensitive patterns start leaking into output.

Related MCP server: chuangsiai-mcp

Install

npm install -g mcp-privacy-shield

Or run on demand with npx:

npx mcp-privacy-shield

Client configuration

Add to your MCP client config (e.g. Claude Desktop, Cursor, or opencode):

{
  "mcpServers": {
    "privacy-shield": {
      "command": "npx",
      "args": ["mcp-privacy-shield"]
    }
  }
}

Tools

Tool

Arguments

Description

sanitize_text

text: string, mask?: boolean

Redact PII → tokens or masks. Returns redacted text, replacement map, triggered categories, and risk.

classify_zone

text: string

Classify into public / internal / restricted / vault with confidence and reasons.

compliance_check

text: string, required_zone?: string

Score 0-100, list violations, and report pass/fail.

shield_catalog

List supported patterns and env config (no secrets exposed).

Environment variables

Variable

Default

Description

MCP_SHIELD_NAMES

0 (off)

Set to 1 to enable capitalized-name pseudonymization ([PERSON_N]). High-confidence only, and off by default to avoid over-redaction.

MCP_SHIELD_ACCOUNT_PATTERN

\b\d{8,17}\b

Regex used to detect bank/routing account numbers.

Detected categories

  • Emails, phone numbers, credit-card numbers, US SSNs, bank/routing account numbers, IPv4 addresses, URLs, and passport/ID patterns.

  • Optionally, full personal names (two consecutive capitalized words) when MCP_SHIELD_NAMES=1.

What this does NOT do

This server is a gate, not a compliance guarantee. Specifically:

  • It is heuristic/deterministic regex matching; it can miss novel or obfuscated PII and can produce false positives.

  • No Luhn validation, no OCR, no cross-referencing against business records.

  • It does not encrypt, store, or transmit anything, and it does not itself enforce legal compliance (GDPR, HIPAA, CCPA, etc.).

  • Before relying on it for regulated data, pair it with a human-reviewed policy and a real data-loss-prevention layer.

Use it as one layer in a defense-in-depth strategy — not as a substitute for professional compliance review.

License & Attribution

MIT — Copyright (c) 2026 Jean Machuca (see LICENSE).

This server implements concepts from the book Vibe Leading The AI: The Corporate Race Against Machines by Jean Machuca (ISBN 9798252505008, © 2026 Jean Machuca). The book is copyrighted commercial material; this repository does not republish its text. Buy the book at https://vibeleading.org or https://a.co/d/04L5YatK. Author website: jeanmachuca.com · Support on GitHub Sponsors: github.com/sponsors/jeanmachuca. See NOTICE.

A
license - permissive license
Not graded
quality - not tested
C
maintenance

Maintenance

Maintainers
Response time
Release cycle
Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

  • A
    license
    Not graded
    quality
    D
    maintenance
    Provides real-time content safety protection for large language models by detecting and preventing risks in both input and output content across multiple dimensions including compliance, ethics, and security.
    MIT
  • A
    license
    Not graded
    quality
    D
    maintenance
    Provides real-time content security for large language models by identifying and intercepting risks across compliance, ethics, and safety dimensions. It enables secure input and output monitoring through a customizable policy engine using an SSE-based interface.
    1
    MIT
  • A
    license
    Not graded
    quality
    C
    maintenance
    Enables AI agents to scan, redact, and govern sensitive data before sending it to external tools through DLP, secure chat, and shadow AI discovery tools.
    AGPL 3.0

View all related MCP servers

Related MCP Connectors

  • Classify data safety before storing or sharing. GDPR, HIPAA, PCI-DSS, CCPA. AI-powered.

  • Responsible-AI guardrails for agents: scoring with policy, injection & PII detection, DPDP.

  • Deterministic trust gate for AI output: leaked-secret, prompt-injection & PII in one call.

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/VibeLeading/mcp-privacy-shield'

If you have feedback or need assistance with the MCP directory API, please join our Discord server