> ## Documentation Index
> Fetch the complete documentation index at: https://docs.orbit.devotel.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Decode the composite verdict

> How the fused SMS-pumping, URL-reputation, Verify fraud, and voice-biometrics signals resolve into one 0-100 score, a band, and an advisory action.

## Decode the composite verdict

Score a destination before you commit a send or a call. Fraud and trust-operations teams wire this endpoint into their pre-send gate; callers also run it ahead of a high-value voice dispatch. The endpoint is read-only and advisory — it sends nothing, routes nothing, and enforces nothing — so you decide how to act on the recommendation it returns.

The composite score folds four detectors into one verdict. The live SMS-pumping / artificial-traffic check and (when you pass `message_body`) the outbound link-reputation scan run on the platform side; your own Verify Fraud Guard and Voice Biometrics scores pass in as `verify_signal` / `voice_biometrics_signal` and are folded into the same number. The composite is the worst present channel score, so a bad link or a flagged destination cannot be washed out by a clean aggregate. Band cutoffs reuse the SMS-pumping scorer's thresholds, so a "high" verdict means the same thing whether the signal came from one channel or the fused whole.

<ResponseExample>
  ```json 200 theme={null}
  {
    "data": {
      "destination": "+14155552671",
      "channel": "sms",
      "score": 82,
      "band": "critical",
      "recommendation": "block",
      "channels": [
        {
          "channel": "sms_pumping",
          "score": 82,
          "present": true
        },
        {
          "channel": "url_reputation",
          "score": 34,
          "present": true
        }
      ]
    },
    "meta": {
      "request_id": "req_risk",
      "timestamp": "2026-08-01T12:00:00.000Z"
    }
  }
  ```
</ResponseExample>

Score bands decode as: **0–24 `low`**, **25–54 `elevated`**, **55–79 `high`**, **80–100 `critical`.** Bands map to actions as `critical → block`, `high → review`, and `elevated | low → allow`. No pass-through scores and no `message_body` returns a clean `score: 0` / `band: "low"` — the empty ask never reads as a threat. The per-channel `channels` array tells you which detector drove the verdict, so the composite is never a black box.