PromptLock Guard for n8n

🛡️ AI-powered security guardrails for your n8n workflows

PromptLock Guard is a community node that adds content analysis and sensitive data detection to your n8n automations. Detect prompt injection attacks, identify PII/PHI patterns, and route content based on risk assessment.

✨ Features

• 🔍 Sensitive Data Detection: Identifies patterns associated with HIPAA, GDPR, and PCI requirements
• 🚦 4-Output Routing: Route workflows based on risk assessment (Allow/Flag/Redact/Block)
• 🔒 Fail-Closed Security: Secure by default with configurable error handling
• ⚡ Real-time Analysis: Fast API integration with configurable timeouts
• 🎯 Flexible Targeting: Support for nested field paths with dot notation
• 📊 Rich Metadata: Detailed analysis results attached to each item

Note: Prompt injection detection is always enabled by default, regardless of which compliance frameworks are selected.

🚀 Installation

Community Nodes (Recommended)

1. In n8n, go to Settings → Community Nodes
2. Click Install a community node
3. Enter: n8n-nodes-promptlock-guard
4. Click Install
5. Restart n8n

npm Installation

# Install globally for n8n
npm install -g n8n-nodes-promptlock-guard

# Or install in your n8n user directory
cd ~/.n8n/
npm install n8n-nodes-promptlock-guard

# Restart n8n

⚙️ Setup

1. Create Credentials

In n8n, create new PromptLock API Key credentials:

• Base URL: https://api.promptlock.io
• API Key: Your PromptLock API key (starts with ps_)
• Header Style: X-API-Key (preferred) or Bearer Token

Get your API key at promptlock.io

2. Add the Node

1. Search for "PromptLock Guard" in the node panel
2. Configure:
   • Text Field: Path to your text data (e.g., text, payload.message)
   • Frameworks: Select detection frameworks (HIPAA, GDPR, PCI)
   • Credentials: Select your PromptLock API Key

3. Wire the Outputs

The node provides four distinct outputs:

• ✅ Allow: Content is safe, proceed normally
• ⚠️ Flag: Content needs review, proceed with caution
• 🔒 Redact: Content has been cleaned, use cleanText field
• 🚫 Block: Content is blocked, do not proceed

📋 Quick Example

Webhook → PromptLock Guard
├─ Allow → Process Normally
├─ Flag → Send to Review Queue
├─ Redact → Process with Clean Text
└─ Block → Return 403 Error

🔧 Configuration Options

Core Settings

• Text Field: Path to analyze (supports dot notation like data.message.text)
• Detection Frameworks: Select HIPAA, GDPR, and/or PCI pattern detection
• Action on High Risk: Override server policy (inherit/flag/redact/block/score)

Advanced Settings

• Write Clean Text To: Field path for redacted content (default: cleanText)
• Attach Metadata Under: Field path for analysis results (default: promptLock)
• On API Error: Error handling strategy (block/flag/allow/throw)
• Request Timeout: API timeout in milliseconds (default: 15000)

📊 Metadata Structure

The node attaches analysis data to each item:

{
  "promptLock": {
    "risk_score": 56,
    "action_taken": "redact",
    "clean_text": "Patient [HIPAA_PERSON_NAME] (SSN: [HIPAA_SSN]) needs treatment",
    "violations": [
      {
        "type": "pii_detection",
        "category": "person_name",
        "confidence": 0.95,
        "position": [8, 18],
        "text": "John Smith",
        "placeholder": "[HIPAA_PERSON_NAME]",
        "compliance_frameworks": ["HIPAA"]
      }
    ],
    "compliance_status": {
      "HIPAA": 2,
      "GDPR": 0,
      "PCI": 0
    },
    "usage": {
      "tokens_analyzed": 8,
      "processing_time_ms": 78
    }
  }
}

🔒 Security Best Practices

• On API Error: Keep as "Block (Fail Closed)" for maximum security
• Action on High Risk: Use "Inherit from Policy" to leverage server-side rules
• Write Clean Text To: Use a separate field to preserve original data

⚠️ Important Note

PromptLock is a security tool that helps detect sensitive data patterns and prompt injection attempts. It is not a compliance certification and does not guarantee regulatory compliance. You remain responsible for your own compliance obligations.

📞 Support

• Website: promptlock.io
• Email: [email protected]

📜 License

MIT License - see LICENSE file for details.

Built by the PromptLock team