Trust center

Trust is useful only when it's scoped, visible, and revocable.

Not blind permission — a clear checkpoint, a passported agent, an enforceable decision, and a path to stop trust when risk changes.

Principles

Four commitments behind every decision.

01

Business control

The business decides which actions need a checkpoint, which agents pass, and which need human approval.

02

Agent-owner accountability

A passport ties the agent to owner, purpose, runtime, scope, and status — never anonymous traffic.

03

Revocation by design

If an agent expires, pauses, or is revoked, the next sensitive decision stops instead of riding on stale trust.

04

Audit without data sprawl

Decision trails show who asked, what was requested, and what was decided — without hoarding raw customer data.

Launch readiness

What's ready, and what still needs verification.

CheckpointLive — businesses can start free with defensive checkpoints.
PassportsLive — $5/month per agent that needs to act online.
Security contactListed — production inbox verification pending before broad launch.
Legal pagesStarter policies — counsel review needed before paid launch.
API accessIssued directly — public pages show the intended model only.
Domain cutoverPreview domain live — gardien.io cutover is a separate step.

Responsible disclosure

Found something? Tell us safely.

Send security reports to security@gardien.io with impact, the affected route or workflow, reproduction steps, timestamps, and safe proof.

Never email live secrets, passwords, private keys, or full customer records. Give us time to triage before public disclosure of an active vulnerability.

The public site is ready for positioning and pilot conversations. Final legal review, inbox verification, and API availability should be confirmed before paid pilots.

Next step

Have a workflow that needs a trust checkpoint?