Connect an AI assistant with the MCP connector
Connect an AI assistant (Claude, or any MCP client) directly to your Northbeams workspace and ask questions about your own data: "what shadow AI showed up this month?", "which tools are marked high-risk?", "are we EU AI Act ready?".
It reads the same data as the REST API at /api/v1/*, exposed as MCP tools.
What you need
- A Northbeams workspace on the Sentinel tier or higher (API access is a Sentinel entitlement).
- An API token. Create one in the dashboard under Settings, then API access. It looks like
nbm_...and is shown once, so copy it somewhere safe.
Connect it
Server URL: https://monitor.northbeams.com/api/mcp
In Claude (Settings, then Connectors, then Add custom connector), add the URL above and provide your token as an Authorization: Bearer nbm_... header. Any MCP client that speaks streamable HTTP works the same way: point it at the URL, send the token on every request.
That is the whole setup. The token identifies your workspace, so the connector can only ever read your own org's data.
Tools
| Tool | What it returns |
|---|---|
list_ai_inventory |
Every AI tool detected in the last 30 days: risk classification, distinct users, event count, first and last seen. |
get_policies |
Your category rules (block, warn, allow), per-tool risk overrides, and Network Control DNS rules. |
get_compliance_status |
EU AI Act readiness attestations and the enforcement date. |
All three are read-only and take no arguments (the token supplies your org).
Quick test
curl -s https://monitor.northbeams.com/api/mcp \
-H "Authorization: Bearer nbm_your_token_here" \
-H "content-type: application/json" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'
You should get back a JSON-RPC result listing the tools.