Masquerade
The Privacy Firewall for LLMs
What is Masquerade?
Masquerade is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to privacy firewall for llms
The Privacy Firewall for LLMs
This server falls under the Security category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- The Privacy Firewall for LLMs
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx masqueradeConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Masquerade
Masquerade is a privacy firewall MCP server that redacts sensitive personal information from PDF documents before they are sent to large language models. It detects entities such as names, emails, dates, and other identifiers using Llama 3.3 70B running on Tinfoil's confidential computing platform, replaces them with pseudonyms, and returns a clean redacted PDF along with a summary of what was masked. This makes it possible to use AI assistants on sensitive documents — legal contracts, medical records, HR files — without leaking personal data to the underlying LLM provider.
Prerequisites
- Python 3.10 to 3.12 installed
- A Tinfoil account and TINFOIL_API_KEY (obtain at tinfoil.sh)
- An MCP-compatible client such as Claude Desktop
- Git installed (for manual pip-based installation)
Obtain a Tinfoil API key
Create a free account at tinfoil.sh and generate an API key from your dashboard. This key authenticates Masquerade's connection to the confidential AI inference platform used for entity detection.
Install Masquerade using the automated setup script
Run the official setup script, which creates a Python virtual environment, installs all dependencies, and configures the Claude Desktop integration automatically.
curl -O https://raw.githubusercontent.com/postralai/masquerade/main/setup.sh && bash setup.shOr install manually into a virtual environment
Create a dedicated Python virtual environment, install Masquerade from GitHub, and then run the configuration helper that updates claude_desktop_config.json.
python3.12 -m venv pdfmcp
source pdfmcp/bin/activate
pip install git+https://github.com/postralai/masquerade@main
python -m masquerade.configure_claudeSet the TINFOIL_API_KEY in your MCP config
The configure_claude script updates claude_desktop_config.json automatically. Verify the entry includes your TINFOIL_API_KEY and the correct Python executable path from the virtual environment.
{
"mcpServers": {
"masquerade": {
"command": "/path/to/pdfmcp/bin/python",
"args": ["-m", "masquerade"],
"env": {
"TINFOIL_API_KEY": "your_tinfoil_api_key"
}
}
}
}Restart Claude Desktop and test with a PDF
Restart Claude Desktop to load the new MCP server. Then provide a PDF file path and ask the AI to redact it. The server will return a redacted PDF and a summary report.
Masquerade Examples
Client configuration
claude_desktop_config.json entry for Masquerade with Tinfoil API key and the virtual environment Python path.
{
"mcpServers": {
"masquerade": {
"command": "/Users/me/pdfmcp/bin/python",
"args": ["-m", "masquerade"],
"env": {
"TINFOIL_API_KEY": "your_tinfoil_api_key"
}
}
}
}Prompts to try
Example prompts for redacting PDFs and reviewing the anonymization results.
- "Redact sensitive information from this PDF: /Users/me/Documents/contract.pdf"
- "Anonymize the personal data in /path/to/medical-record.pdf and show me a summary of what was removed."
- "Process this HR document and replace all employee names with pseudonyms: /tmp/hr-report.pdf"
- "How many entities were redacted on each page of the last PDF you processed?"Troubleshooting Masquerade
TINFOIL_API_KEY not recognized or authentication fails
Confirm your key is valid by logging into tinfoil.sh and regenerating it if needed. Ensure the key is set in the env block of your MCP config, not as a shell environment variable, since MCP servers run in isolated processes.
Python version error on startup
Masquerade requires Python 3.10 to 3.12. Check your version with python3 --version and install a compatible version if needed. Use the exact Python executable from the virtual environment in your MCP config command path.
Redacted PDF is not created or path is missing
Ensure the PDF path you provide is an absolute path and the file exists and is readable. The server processes files locally, so relative paths may not resolve correctly depending on the MCP client's working directory.
Frequently Asked Questions about Masquerade
What is Masquerade?
Masquerade is a Model Context Protocol (MCP) server that privacy firewall for llms It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Masquerade?
Follow the installation instructions on the Masquerade GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Masquerade?
Masquerade works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Masquerade free to use?
Yes, Masquerade is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Masquerade Alternatives — Similar Security Servers
Looking for alternatives to Masquerade? Here are other popular security servers you can use with Claude, Cursor, and VS Code.
Casdoor
★ 13.6kAn open-source Agent-first Identity and Access Management (IAM) /LLM MCP & agent gateway and auth server with web UI supporting OpenClaw, MCP, OAuth, OIDC, SAML, CAS, LDAP, SCIM, WebAuthn, TOTP, MFA, Face ID, Google Workspace, Azure AD
ghidraMCP
★ 9.0kAn Model Context Protocol server that enables LLMs to autonomously reverse engineer applications by exposing Ghidra's decompilation and analysis tools. It allows AI agents to list code structures, rename methods, and analyze binaries directly through
HexStrike AI
★ 8.9kHexStrike AI MCP Agents is an advanced MCP server that lets AI agents (Claude, GPT, Copilot, etc.) autonomously run 150+ cybersecurity tools for automated pentesting, vulnerability discovery, bug bounty automation, and security research. Seamlessly b
IDA Pro MCP
★ 8.7kEnables AI-assisted reverse engineering in IDA Pro by providing tools to analyze binaries, decompile functions, manage comments, search patterns, and interact with the IDA database through natural language.
Anthropic Cybersecurity Skills
★ 6.6k754 structured cybersecurity skills for AI agents · Mapped to 5 frameworks: MITRE ATT&CK, NIST CSF 2.0, MITRE ATLAS, D3FEND & NIST AI RMF · agentskills.io standard · Works with Claude Code, GitHub Copilot, Codex CLI, Cursor, Gemini CLI & 20+ platform
Hooker
★ 5.1k🔥🔥 hooker is a Frida-based reverse engineering toolkit for Android. It offers a user-friendly CLI, universal scripts, auto hook generation, memory roaming to detect activities/services, one-click SOCKS5 proxy setup, Frida JustTrustMe, and BoringSSL u
Browse More Security MCP Servers
Explore all security servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Masquerade in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Masquerade?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.