API Integration
The most flexible way to integrate Signal Canary AI DLP is through our REST API. This works with any AI application, chatbot, or custom integration.
Getting Your API Key
1 Go to Settings
Navigate to Dashboard > Settings > API Keys.
2 Generate API Key
Click Generate New Key and copy the key securely.
Scanning Content
Send content to the DLP scan endpoint before passing it to your AI:
POST /api/v1/dlp/scan
Content-Type: application/json
X-API-Key: your-api-key
{
"content": "User prompt to scan",
"policy_id": "your-policy-id"
}Response Format
{
"allowed": true,
"detections": [],
"sanitized_content": "User prompt to scan",
"action_taken": "none"
}
// Or when sensitive data is found:
{
"allowed": false,
"detections": [
{
"pattern": "credit_card",
"match": "4111-****-****-1111",
"position": 45
}
],
"sanitized_content": null,
"action_taken": "block"
}Example: Python Integration
import requests
def scan_before_ai(user_prompt):
response = requests.post(
'https://signalcanary.net/api/v1/dlp/scan',
headers={'X-API-Key': 'your-api-key'},
json={
'content': user_prompt,
'policy_id': 'default'
}
)
result = response.json()
if not result['allowed']:
return None, result['detections']
return result.get('sanitized_content', user_prompt), []
# Usage
safe_prompt, violations = scan_before_ai(user_input)
if violations:
print("Sensitive data detected:", violations)
else:
# Safe to send to AI
ai_response = call_your_ai(safe_prompt)Proxy Mode
For simpler integration, route your AI API calls through Signal Canary's proxy. We'll scan requests automatically.
How It Works
1 Change Your Base URL
Instead of calling the AI provider directly, call through Signal Canary:
# Instead of:
https://api.openai.com/v1/chat/completions
# Use:
https://proxy.signalcanary.net/openai/v1/chat/completions2 Add Your Credentials
Include both your AI provider key and Signal Canary key:
Authorization: Bearer your-openai-key
X-Signal-Canary-Key: your-signal-canary-key3 Make Requests Normally
Your requests work exactly as before, but are now scanned for sensitive data.
Browser Extension
For web-based AI tools like ChatGPT, Claude, or Gemini, our browser extension scans prompts before they're sent.
Installation
1 Install Extension
Download from Chrome Web Store or Firefox Add-ons (coming soon).
2 Connect Your Account
Click the extension icon and log in with your Signal Canary credentials.
3 Configure Sites
Choose which AI sites to protect (ChatGPT, Claude, etc.).
How It Works
The extension intercepts form submissions on AI chat interfaces, scans the content against your DLP policies, and either allows, warns, or blocks based on your settings.
Webhooks & Alerts
Get notified instantly when DLP policies detect sensitive data.
Setting Up Webhooks
1 Go to Integrations
Navigate to Dashboard > Settings > Integrations.
2 Add Webhook URL
Enter your webhook endpoint URL where you want to receive events.
3 Select Events
Choose which events trigger webhooks (detections, blocks, etc.).
Webhook Payload
{
"event": "dlp.detection",
"timestamp": "2024-01-15T10:30:00Z",
"policy": {
"id": "policy-123",
"name": "Block PII"
},
"detections": [
{
"pattern": "ssn",
"action": "blocked"
}
],
"metadata": {
"source": "api",
"ip": "203.0.113.1"
}
}Integration Examples
- Slack - Get alerts in a security channel
- PagerDuty - Trigger incidents for critical blocks
- SIEM - Send events to Splunk, Datadog, or similar
- Custom - Build your own response workflows
Choose the integration method that best fits your workflow. API integration offers the most control, while the browser extension is the quickest to deploy.