Terms of Service
These terms cover your use of DidWork at didwork.sh, its API, its SDKs, its MCP server and its editor plugins.
Last updated 17 September 2026.
The service
DidWork takes a claim that an outcome happened, checks it read-only
against the system that can prove it, and returns a verdict of
VERIFIED, FAILED
or UNKNOWN, with the evidence it gathered.
It is a verification service, not a monitoring or incident-response service. It answers when asked, and on the interval you set for a watch. It does not guarantee it will notice anything you have not asked it to check.
Your account
You need an account to get an API key. You are responsible for keeping your keys secret and for what is done with them: a key is a bearer secret, so anything holding it can verify as you and spend your allowance. Revoke a key in the console if it leaks.
One person or organisation per account. Do not share a single account across parties who should not see each other's verifications.
Your credentials and data
You keep ownership of everything you send us and everything you connect. Granting DidWork a read-only credential does not transfer any right in the account behind it.
We use your data to run the service you asked for. We do not sell it, and we do not use the contents of your verifications or your connected accounts to train models. What we store, who processes it, and how long it lasts is in the Privacy Policy.
What a verdict is, and is not
A verdict is a statement about the evidence DidWork could gather at the
moment it asked, from the system you pointed it at. It is deliberately
conservative: when evidence cannot be reached, you get
UNKNOWN rather than a guess, and
UNKNOWN never rounds up to
VERIFIED.
It is not a warranty that the underlying outcome is correct, lawful, or suitable for anything. A provider can report state that is itself wrong, and a verified claim only means the provider's record matched what you asserted. Do not use a verdict as the only control on an action that cannot be undone.
Acceptable use
Do not use DidWork to:
- verify against systems you are not authorised to read;
- point
http.okat hosts you do not own or have permission to probe, or use watches to generate load against a third party; - work around the rate limits, the free allowance, or the keyless endpoint's restrictions;
- resell raw verification capacity as your own verification service.
Requests to internal and private network addresses are blocked, and that protection is not something to route around.
Billing
Free includes 1,500 verifications each month, one watch at hourly
cadence, one API key, and 7 days of evidence. No card. Past the
allowance, requests return 402 and watches
pause; nothing is deleted for that reason, and everything resumes when
you subscribe.
Pro is $49 per month and includes 2,500 verifications, then $0.01 for
each one beyond that. One verification is one claim checked once, so a
did.verify() call and a single watch run cost
the same. Requests refused at the free cap are never counted or billed.
Payments are handled by Stripe. We do not see or store your card number.
Cancelling
Cancel Pro at any time from the billing portal. You keep Pro until the end of the period you have paid for, then the account returns to Free. Evidence older than 7 days is deleted that day, so export anything you still need first.
Deleting your account is separate and immediate, and it is described in the Privacy Policy.
Changes
We will update these terms as the product changes. The date at the top says when they last moved.