Gate every AI agent inference against regulatory compliance frameworks. One API call returns a verdict, five auditable checks, and a cryptographic receipt.
Hardware production authentication
Pre-inference compliance certification
Multi-media forensic analysis
All API requests require a bearer token. Get your key from the dashboard.
Authorization: Bearer nxc_your_api_key_here
Unauthenticated requests use a demo tier (10 req/min). Authenticated requests use your plan limits.
| Field | Type | Description |
|---|---|---|
| input | string | The agent output to certify required |
| agentId | string | Identifier for the calling agent required |
| framework | string | Regulatory framework to certify against required |
curl -X POST https://www.debacconexus.ai/v1/compliance/certify \
-H "Authorization: Bearer nxc_your_key" \
-H "Content-Type: application/json" \
-d '{
"input": "Your agent output here",
"agentId": "my-agent-v1",
"framework": "EU_AI_ACT"
}'
{
"success": true,
"certified": true,
"verdict": "CERTIFIED",
"reason": "Specific finding for this output",
"framework": "EU_AI_ACT",
"checks": {
"transparency": { "pass": true, "finding": "..." },
"humanOversight": { "pass": true, "finding": "..." },
"accuracy": { "pass": true, "finding": "..." },
"dataGovernance": { "pass": true, "finding": "..." },
"riskManagement": { "pass": true, "finding": "..." }
},
"receiptId": "NXC-MTN64BOV",
"timestamp": "2026-09-04T16:28:32.053Z",
"signature": "NXC-MTN64BOV|CLOUD|EU_AI_ACT|35F46CC8"
}
Every request evaluates agent output against five checks derived from the selected framework.
Does the output disclose its AI origin?
Can a human intervene before action is taken?
Is the output factually grounded?
Does the output respect data handling rules?
Are risks identified and mitigated?
Output passes all checks. Safe for production delivery.
Output fails one or more checks. Blocked before reaching the end user.
Certify up to 50 agent outputs in a single request. Each item receives its own verdict and receipt.
curl -X POST https://www.debacconexus.ai/v1/compliance/batch \
-H "Authorization: Bearer nxc_your_key" \
-H "Content-Type: application/json" \
-d '{
"framework": "NIST_AI_RMF",
"items": [
{ "input": "Agent output 1", "agentId": "agent-a" },
{ "input": "Agent output 2", "agentId": "agent-b" }
]
}'
Public endpoint — no authentication required. Any party can independently verify a certification receipt.
curl https://www.debacconexus.ai/v1/receipts/NXC-MTN64BOV
Returns the full receipt including verdict, checks, signature, and verification timestamp.
| EU_AI_ACT | European Union Artificial Intelligence Act |
| NIST_AI_RMF | NIST AI Risk Management Framework |
| ISO_42001 | ISO/IEC 42001 AI Management System |
Enterprise customers can define custom compliance frameworks.
Every request passes through three independent certification layers. One receipt covers all three.
Forensic verification that content was produced by real hardware
Agent output gated against the selected regulatory framework
Text, image, voice, and video scanned for AI generation markers
No publicly documented product combines all three layers in a single receipted system.