Agent ToolkitFreeNothing leaves your browser

Know what your AI agents are really doing.

Five free tools for teams running AI agents. Score your risk, put a price on the mistakes nobody catches, and get every agent on paper. Everything runs in this page; nothing is uploaded or sent anywhere.

Measure your risk

2 tools

Agent Risk Scorecard

Score your agent risk

Answer a few questions about how your agents are set up. Get a risk score, your three biggest gaps and how to close them. Your cost exposure is then worked out from your answers.

About 3 minutesInstant result

Exposure Calculator

Estimate what agent mistakes cost you

Enter your agent volume and the typical cost of an error. See the estimated yearly cost of the mistakes your agents make that nobody catches.

About 1 minuteEvery assumption editable
Estimated annual exposure · example
$6,480 a year
50,000 decisions a month, caught only by complaints
Open the calculator →

Put your agents in order

3 tools

Agent Passport

Give an agent an identity

Its job, what it may and must never do, who approves what, and who is accountable. Copy it into your docs.

Create a passport →

Agent Register

List every agent you run

One table of all your agents, a risk map across them, and an export for Excel, Sheets or Notion.

Open the register →

Failure Simulator

Watch a mistake unfold

Pick a realistic agent failure and see how it spreads, when you would find out, and what it costs.

Run a simulation →
Private by designNo sign-up, no tracking, no upload. Your answers stay on this page; the register is saved only in this browser.
Plain-language resultsEvery score shows its reasoning, and every assumption in the estimates can be changed.
An indication, not a verdictThese tools estimate risk from what you tell them. WOWN checks what your agents actually do, decision by decision.

← All tools
0risk score / 100

Risk by area

Your three biggest gaps

Where a costly agent mistake is most likely to go unnoticed.

Exposure Calculator

What could agent mistakes cost you?

Enter your own figures, or keep our defaults. Every assumption is yours to change.

Every answer your agents give or action they take.
Decisions that move money, place orders or promise customers something.
Among decisions with real-world effect. Default: 1 in 5,000.
Refunds, rework, write-offs, staff time, goodwill.
Only uncaught errors count as exposure.
Enter costs in the currency you choose.
Estimated annual exposure
$0 a year
Caught todayUnnoticed

Typical case your inputs

Cautious case half the errors

How this is calculated
  • Costly errors a year = decisions per month × 12 × share with real-world effect ÷ error frequency.
  • Unnoticed errors = costly errors × share not caught today.
  • Annual exposure = unnoticed errors × average cost of one error.
  • Cautious case: the same calculation with half as many errors.
An estimate based on your inputs and the stated assumptions, not a guarantee of savings.

What next

All tools

Agent Passport

Give your agent an identity document

Write down who your agent is, what it may do, what it must never do and who answers for it. You get a passport you can copy into your documentation, onboarding or risk files.

What it is allowed to do
Which of these need a human to approve first
Only actions it is allowed to do are shown.

Agent Register

Every agent you run, in one place

Keep one list of the agents you run and see which ones carry the most risk. It is often the first thing a board asks for. Your register is saved in this browser only.

AgentHighest actionHuman approvalCheckedDecisions / monthOwnerRiskRemove

Your agent risk map

Failure Simulator

Watch one agent mistake unfold

Pick an agent and a realistic failure. See how it spreads, when you would find out, and what it costs, depending on how your agent is checked today.

01Your agent

02What goes wrong

03Your situation

Pre-filled for this scenario. Change it to yours.
Found after
Incidents by then
0
Estimated cost
0

Checked today vs. checked in real time

How you check today
Every decision checked

An illustrative scenario. Frequencies and costs are assumptions you can change; real outcomes depend on your agent, your data and your customers.