Free tools
Five tools covering the decisions an agent project actually turns on — whether to build, what to specify, why it won't run unattended, and whether the vendor in front of you has done this before. No email gate on any of them.
Get my free blueprintEvery one of these exists because we needed it ourselves or got asked the same question enough times to write it down. None of them gate the result behind an email — you get the answer, and if you want a human to look at it, that is your choice rather than the price of entry.
They are ordered below by where you are in the project, because using the production diagnostic before you have built anything is as useless as using the readiness checklist after you have already shipped.
Is this workflow worth automating at all?
Eight yes/no questions on scope, data access, volume and ownership. Tells you what to fix before an agent is even the right conversation.
Open toolWhat is this repetitive task actually costing us?
Put your own numbers on one workflow and see the year laid out. Uses your figures rather than a heroic number picked to sound good.
Open toolAn email gate on a calculator buys you a list of people who wanted a number and now resent you slightly. We would rather the tool be good enough that some of the people who use it choose to get in touch — and that the ones who do arrive already knowing what they need.
Each tool does offer to send its result to us, with the full working attached, if you want a second opinion. That is a button, not a wall.
It depends where you are. Comparing vendors: the vendor scorecard. Deciding whether a workflow is worth automating: the readiness checklist, then the ROI calculator to size it. About to build: the spec generator. Already built something that will not run unattended: the production readiness diagnostic.
No. Every tool computes its result in your browser and shows it immediately. Each one then offers to send the result to us with the full working attached if you want a second opinion, but that is optional and the answer is already on screen.
The checklist is a pre-build question: is this workflow repetitive, well-scoped and high-volume enough to justify an agent. The diagnostic assumes you already have a working agent and asks which of the five failure classes is stopping it from running unattended. Different stage, different questions.
That is the point of publishing it. Our answers to the four deal-breaker criteria are in our case studies: a documented failure taxonomy with frequencies, agents running unattended with restart counts, spend ceilings enforced in code, and self-hosted deployment by default.
Tell us what eats your week. We’ll come back with a concrete plan for the first agent to build and what it’ll automate — before you spend anything.
Goes straight to hello@iamagentman.com — we read every message ourselves. Prefer to answer three questions instead?Build your blueprint.
No retainer to start · reply within 24 hours