Getting Started with Berenta

Berenta is a secure AI agent platform. Agents on Berenta are autonomous programs that can call external APIs, process data, and take actions — all within a policy boundary you define and control.

How it works

1

Build or browse an agent

Use the Forge to configure a new agent from scratch, or browse the Marketplace to deploy a pre-built agent to your account.

2

Define the policy

Every agent has a policy — a structured document that defines its system prompt, the tools it can use, the external domains it may contact, and any secrets it requires. You own this policy and can update it at any time.

3

Run and monitor

Chat with your agent directly from the dashboard. When an agent attempts a high-risk action, you'll receive a real-time approval request before execution proceeds. All activity is logged to your audit trail.

4

Publish to the Marketplace (optional)

If you've built an agent others would find useful, you can submit it for review and list it on the Marketplace. Buyers receive an immutable copy — your original is never modified.

Key concepts

Agent

An autonomous program with a signed identity, a policy, and a set of permitted tools.

Policy

The configuration document that governs what an agent can and cannot do.

Forge

The interface for building and configuring agents.

Marketplace

The catalog of pre-built, security-reviewed agents available for deployment.

Need help?

For support, contact support@berenta.com. Review our Terms of Service and Privacy Policy.

Browse the Marketplace