Quick Start
Get from login to your first governed agent call in under 10 minutes.
Prerequisites
- A Behavry account (contact your administrator for an invitation link)
- An AI agent or MCP-compatible client (Claude Code, Cursor, ChatGPT, etc.)
Step 1 — Log In
Navigate to your Behavry dashboard URL and sign in with your credentials.
After login you'll land on the Activity page, which shows a live feed of all agent actions. It will be empty until you enroll an agent.
Step 2 — Get Your Enrollment Token
- Go to Settings → Agents
- Click New Enrollment Token
- Copy the token — you'll use it once to register your first agent
Enrollment tokens are single-use. Each agent gets its own identity after registration.
Step 3 — Connect Your Agent
Follow the guide for your AI client:
Or see the full list in Connecting Agents →
Step 4 — Make a Test Call
After connecting, trigger any action from your agent (e.g., ask it to read a file or call an API). Within seconds you should see the event appear in the Activity feed with:
- Agent identity confirmed ✓
- Policy evaluated ✓
- Action allowed, escalated, or blocked based on your policy
Step 5 — Set Your First Policy
By default, Behavry runs in report-only mode — it logs and evaluates every action but doesn't block anything. Once you've confirmed your agent is connected:
- Go to Policies
- Click New Policy
- Write your first rule or use a built-in template
See Write Your First Policy → for a guided walkthrough.
Need Help?
- Troubleshooting connection issues → Troubleshooting
- Self-hosted deployment → Docker Compose
- Contact support → support@behavry.ai