V2 API Reference
The Sentinel V2 API is a deterministic, multi-tenant evaluation endpoint. It does not perform active synchronous background profiling, guaranteeing <10ms response times.
1. Trust Evaluation Endpoint
Evaluate an incoming IP address against your specific tenant's DSL policies and the internal behavioral hash map.
Endpoint: POST /v2/evaluate
Headers:
Authorization: Bearer <YOUR_API_KEY>Content-Type: application/jsonx-sentinel-trust: <JWT>(Optional. Required to bypass a block after a challenge is successfully solved.)
Request Body
json
{
"target": "105.113.18.224"
}JSON Response Schema
Success Example (Human Evaluator):
json
{
"status": "success",
"tenant": {
"api_key": "VALID",
"policy_engaged": "BALANCED"
},
"decision": {
"verdict": "ALLOW",
"score": 85,
"signals": [
{
"id": "RESIDENTIAL_IP",
"weight": 10,
"label": "Residential/Organic Network"
}
],
"latency_ms": 4
}
}Challenge Example (Bot Suspicion):
json
{
"status": "success",
"tenant": {
"api_key": "VALID",
"policy_engaged": "BALANCED"
},
"decision": {
"verdict": "CHALLENGE",
"score": -20,
"signals": [
{
"id": "SCANNER_PATTERN",
"weight": -50,
"label": "Fast Enumeration Detected"
}
],
"action_required": "SOLVE_CAPTCHA",
"latency_ms": 5
}
}2. Verdict Definitions
| Verdict | Meaning | Required Action |
|---|---|---|
ALLOW | The request is structurally and behaviorally human. | Forward the request. |
BLOCK | The request aggressively triggered a critical threat parameter. | Deny request with a 403 or 401. |
CHALLENGE | The request hit an ambiguous score threshold. | Serve the SOLVE_CAPTCHA Silent Widget, then re-evaluate. |
3. Asynchronous Data
V2 prioritizes speed over synchronized data fetching. If an IP lacks Deep Forensic Data (ASN/Carrier) during the first request, the Engine triggers a parallel cold cache retrieval. Subsequent requests from that IP within 24 hours will include the deeply analyzed data.