Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.svantic.com/llms.txt

Use this file to discover all available pages before exploring further.

​
Security

Svantic is designed for regulated industries where data sovereignty, auditability, and access control are non-negotiable. This page describes the security architecture from a user’s perspective.

​
Authentication

Every agent authenticates with Svantic using API credentials (Client ID and Client Secret). You obtain these from the dashboard under Settings → API Keys. The SDK handles authentication automatically — provide your credentials and the SDK exchanges them for a short-lived token, then uses that token for all subsequent API calls. 
const mesh = new MeshConnector(agent, {
	svantic_url: 'https://api.svantic.com',
	client_id: process.env.SVANTIC_CLIENT_ID!,
	client_secret: process.env.SVANTIC_CLIENT_SECRET!,
});
await mesh.connect();
No credentials are sent after the initial token exchange. The token expires periodically and the SDK refreshes it automatically.

​
Organization Isolation

Each Svantic account is cryptographically isolated:
  • Agents registered under your account are invisible to other accounts
  • Sessions, messages, and knowledge are scoped to your account
  • Routing decisions never cross account boundaries
  • Even on shared infrastructure, your data is isolated at the authentication layer — not by application logic

​
Data Sovereignty

Svantic never touches your data. Capabilities execute on your infrastructure — your agents run on your machines, access your databases, and use your credentials locally. Svantic only sees:
  • Capability names and descriptions (from your agent card)
  • Structured results your agent chooses to return
  • Session metadata (who’s involved, what state things are in)
Your raw data, credentials, file contents, and internal state never leave your environment.

​
Per-Invocation Authentication

Every capability invocation is independently authenticated — not just the connection. Each dispatch carries a short-lived, cryptographically signed token that your agent’s SDK verifies before executing any handler. This means:
  • Replay attacks are blocked (tokens expire in seconds)
  • Each invocation is bound to a specific agent instance
  • A compromised callback URL cannot be used to forge requests
The SDK handles all verification automatically. You don’t write any auth code in your handlers.

​
Safety Layers

Svantic applies multiple independent safety layers to every operation:
LayerWhat It Does
AuthenticationVerifies agent identity before any operation
Invocation AuthCryptographically authenticates every capability call
Registration PolicyControls which agent types can operate (open, allow-list, or audit mode)
ToolGuardGates tool invocations — file access, commands, resource budgets, bulk operations
FlowGuardPrevents runaway execution — step limits, timeouts, cycle detection
Approval WorkflowsRequires human sign-off for sensitive operations
Zero-Knowledge CredentialsCredential values never leave the client machine
Each layer operates independently. Compromising one does not compromise the others.

​
Zero-Knowledge Credentials

When an agent needs to fill in a login form or provide an API key, Svantic uses a zero-knowledge model:
  1. The agent identifies the form field and provides a credential key
  2. The client tool retrieves the actual value from local secure storage
  3. The value is used locally — it is never transmitted to Svantic
Passwords, API keys, and tokens never appear in Svantic’s logs, context, or storage.

​
Compliance

Svantic’s architecture supports compliance with common regulatory frameworks:
  • SOC 2: Full audit trails, access controls, encryption at rest
  • HIPAA: Data never leaves the processing environment
  • PCI-DSS: Zero-knowledge credential handling
  • GDPR: Organization-scoped data isolation, right to deletion

​
Getting Credentials

Sign up at app.svantic.com and navigate to Settings → API Keys → Create Key. You’ll receive a Client ID and Client Secret. Store these securely — the secret is shown only once. See the Getting Started guide for setup instructions.