See Waxell, end to end.

Observe SDK

See & govern every run

One line gives you full traces and cost — then governs every run against 50+ policy categories mapped to OWASP, NIST, ISO 42001 and the EU AI Act.

Runtime

Author, risk-check, run

Declare an agent and push it — get a framework-mapped risk report, a versioned registry entry (any framework), and guards enforced from the first run.

MCP Gateway

Govern every tool

One governed surface in front of every MCP server — redaction, approvals, rug-pull protection, and audit on every call.

Connect

Where your team & agents work

The shared workspace where people and agents do the work — tasks, versioned files, playbooks, and cross-team content hand-offs (dev docs → marketing).

Endpoint

Find & govern shadow AI

A native Mac & Windows agent that finds every AI tool on your devices, measures the governed-vs-ungoverned gap, and blocks what you can’t govern — in under a minute.