Standalone Scan API - Krixai

Use this if you want to scan text without proxying to an LLM. Useful for scanning user inputs before your own processing logic, or for scanning outputs from models you call directly.

POST https://api.krixaisecurity.com/v1/scan

Request Headers

Header	Required	Description
`X-Krixai-Key`	Yes	Your Krixai API key: `kx-live-...`
`Content-Type`	Yes	`application/json`

Request Body

{
  "input": "Text to scan for threats...",
  "checks": ["prompt_injection", "jailbreak", "pii"],
  "sensitivity": "medium"
}

Response

{
  "request_id": "kx-req-def456",
  "status": "flagged",
  "scan_time_ms": 9,
  "detections": [
    {
      "category": "pii",
      "sub_type": "email_address",
      "confidence": 1.0,
      "matched_text": "john.doe@company.com",
      "position": { "start": 45, "end": 67 }
    }
  ],
  "sanitized_input": "Text to scan... contact [REDACTED] for details."
}

⌘I