
Redact MCP
A privacy layer for Claude Code pentest workflows. Automatically detects PII (IPs, emails, API keys, person names, orgs, locations) using regex and NER, swaps them for deterministic fake values, and maintains a bidirectional mapping table. The redact_proxy_request tool deobfuscates outbound HTTP requests, makes the real call to target systems, then obfuscates the response before Claude sees it. Includes redact_read_file for sanitizing input files and redact_export for reversing obfuscation in final deliverables. Optional JSONL audit logging captures every transformation with full before/after text. Useful when you need Claude to analyze scan results or API responses without sending client infrastructure details to an LLM provider.
- GitHub stars
- 6
- tools
- 8
- transport
- STDIO
- env vars
- 2
- last updated
- Mar 23, 2026
Tools
8 toolsPublic tool metadata for what this MCP can expose to an agent.
redact_obfuscateAuto-detect and replace all PII in text (regex + NER)
Auto-detect and replace all PII in text (regex + NER)
No parameter schema in public metadata yet.
redact_deobfuscateReverse all replacements back to real values
Reverse all replacements back to real values
No parameter schema in public metadata yet.
redact_proxy_requestHTTP proxy — deobfuscates request, makes real call, obfuscates response
HTTP proxy — deobfuscates request, makes real call, obfuscates response
No parameter schema in public metadata yet.
redact_read_fileRead a file and return obfuscated content
Read a file and return obfuscated content
No parameter schema in public metadata yet.
redact_add_mappingManually add a real-to-fake mapping
Manually add a real-to-fake mapping
No parameter schema in public metadata yet.
redact_remove_mappingRemove a mapping (fix false positives)
Remove a mapping (fix false positives)
No parameter schema in public metadata yet.
redact_show_mappingsShow all current mappings grouped by type
Show all current mappings grouped by type
No parameter schema in public metadata yet.
redact_audit_logView recent audit log entries (requires REDACT_AUDIT_LOG=true)
View recent audit log entries (requires REDACT_AUDIT_LOG=true)
No parameter schema in public metadata yet.
Sponsors
Configuration
REDACT_AUDIT_LOGEnable JSONL audit trail for compliance tracking
REDACT_DATA_DIRDirectory for mappings and audit log storage
Readme
Redact MCP — Automatic PII Obfuscation for Claude Code
An MCP server (and Claude Code plugin) that automatically detects and obfuscates sensitive data before Claude ever sees it. Uses regex pattern matching and AI-powered Named Entity Recognition (NER) to catch IPs, hostnames, emails, API keys, person names, organization names, locations, private keys, connection strings, and more.
Built for penetration testers who need Claude's analysis capabilities without exposing client data to a third party.
How It Works
Raw data with real PII ──► Regex + NER detection ──► Claude sees only fake values
│
Final report for client ◄── Real values restored ◄── /redact:export
The server maintains a bidirectional mapping table. Every sensitive value gets a consistent, deterministic fake replacement that persists across the entire session:
Details
- Repository
- github.com/r3352/redact-mcp
- Website
- github.com/r3352/redact-mcp
- Category
- Data & Analytics
- Package
@mattzam/redact-mcp- Transport
- STDIO
- Registry
- MCP registry, active