Free Checklist · 14 Pages · PDF
Most agent projects stall when a second team wants in and nobody can say who owns what, what gets tested, what it costs, or who can switch it off. These 33 yes-or-no checks ask exactly those questions, and tell you which one to fix first.
Every check is traced to Microsoft documentation, with the page and its update date, or marked as our own recommendation. We label preview features as preview. See how we verify what we publish.
Prefer to read first? The checks themselves are free on the web, no email needed: read the web version. The PDF adds the evidence prompts, score sheet and the templates.
Tell us a little about your role and the download unlocks on this page. No confirmation link to chase.
What's inside
Four guardrails (an evaluation gate, identity on every model call, spend tied to what shipped, a person where it matters) and a foundation check. Each one links to Microsoft documentation, or says it is our own recommendation.
Each check names the evidence that proves it, so “we think so” scores as partly, not yes.
Find the one guardrail to fix first, and put one accountable name against each decision.
What an engagement with us covers for each gap, and how it starts.
What to ask whoever builds your agents, including us, before you sign.
A template for explaining one agent, or the whole platform, to leadership.
Not sure where to start?
Six questions, five minutes, no email. It maps your result to Microsoft's own Agentic AI adoption maturity model, then points you to the right next read.
Take the Readiness ScorecardLast updated: