MAXIMUM 3 RUNNING AT A TIME

The 10-Day Agent Governance Assessment

We deploy SentnelOps against one of your environments — in shadow mode, in your VPC — and show you exactly what your AI agents can do, what they actually did, and what should never have been possible. You keep the report either way.

[ WHAT YOU GET, DAY BY DAY ]

DAY 1

Deployment

White-glove, screen-share, shadow mode: nothing is blocked, nothing breaks.

DAY 2

Agent inventory

Every agent, every MCP server, every tool it can reach.

DAY 4

Risk findings

Unrestricted actions, reachable secrets, destructive capabilities — with your Agent Blast Radius Score.

DAY 6

Policy recommendations

Two or three least-privilege policies, drafted with your team.

DAY 8

Enforcement simulation

Policies flipped from shadow to enforce on one scope.

DAY 10

Governance report

Auditor-ready. Then one question: do you want to keep enforcement running?

[ IS THIS FOR YOU? ]

You're running AI agents — Claude Code, Cursor, or custom — against real systems: GitHub, AWS, databases, internal APIs. And someone (security, an auditor, you at 2 a.m.) has started asking what those agents can actually do. You can name a technical owner and give us one environment. That's it.

What it costs: the assessment is free for qualified teams. We run a maximum of three at a time.

[ REQUEST YOUR SLOT ]

e.g. Claude Code, Cursor, custom agents
e.g. GitHub, AWS, Postgres, internal APIs

Prefer to start smaller? The Scout tier is free — govern your first MCP server in 10 minutes.