Integration Quick Start

Connect Sentinel SCA to your own agents, automation, IoT systems, or internal tools. Sentinel becomes the checkpoint your system asks before executing sensitive actions.

1. Create Agent

2. Keep Private Key

Your system keeps the Ed25519 private key. Sentinel uses the public identity to verify requests.

3. Ask Sentinel

Before execution, send a signed proposed action to /analyze.

4. Enforce Decision

Your system executes only when Sentinel returns an allowed decision.

Governance Flow

Customer app or AI agent
  -> signs proposed action
  -> POST /analyze
  -> receives allow, deny, or review
  -> executes only if allowed
  -> evidence appears in dashboard

Example Request Shape

{
  "agent_id": "agent_xxx",
  "command": "{\"type\":\"read_url\",\"target\":\"https://example.com\"}",
  "timestamp": "2026-06-08T12:00:00Z",
  "signature": "base64-ed25519-signature"
}

First Integration Checklist

Log into the customer dashboard.
Create or enroll an Ed25519-governed agent.
Copy the agent_id into your application configuration.
Sign proposed actions with the agent private key.
Call /analyze before executing the action.
Block execution when Sentinel returns deny or review.
Confirm the decision appears in dashboard audit history.
Export evidence after the first successful test.

Paddle activates paid package limits. Sentinel handles account access, agents, governance decisions, audit trails, and evidence.

Open Customer Dashboard

One SDK, One API, Package-Enforced Layers

Integrate once. Sentinel applies governance layers automatically based on agent identity, tenant policy, and the customer package.

Developer: same API path with developer limits.
Startup: same API path with Startup agent and usage limits.
Business: same API path with expanded team, audit, and evidence workflows.
Enterprise: same API path with custom limits, support, retention, and deployment review.

The customer app should not implement each Sentinel layer separately. It calls Sentinel before execution; Sentinel enforces identity, replay protection, policy, tenant isolation, package entitlement, and audit evidence.