Skip to content
Agent Mandate

AI agent action authorization API

Authorize AI agent actions against a signed mandate

Before a consequential action runs, verify it against the authority, limits, expiry, and approval requirements captured in a signed mandate.

Run the keyless demo Create a sandbox key
Customer-supplied input No key for the demo No card for a sandbox account

Exact use case

An agent proposes a payment, procurement, account, or data action whose authority must be checked before execution.

Input you provide

A supplied mandate and proposed action, including resource, amount, currency, expiry, limits, and approval rules.

Output you receive

Allowed, denied, or needs-human, plus violations, the deciding clause, policy context, and a durable receipt.

What the API does

Answers whether the proposed action fits the supplied signed mandate and policy at verification time.

Authenticated first value

After mandate issuance, a successful authenticated POST /v1/verify that returns a decision and receipt.

OpenAPI-grounded keyless demo

Verify the action

Edit the supplied example, send it to /v1/demo/verify, and inspect the service’s unedited JSON response. The demo stores nothing and consumes no account quota.

Request body Ready

This sample matches the published request contract for /v1/demo/verify. Do not paste secrets or personal data.

Response
Run the demo to see the API response.

A successful response demonstrates the bounded classification or calculation shown here. It is not evidence of a deployed connector, autonomous external action, customer outcome, or commercial performance.

Capability boundary

What this page does not promise

It does not authenticate the agent identity, replace OAuth, authorize MCP transport, or enforce the decision outside the API response.

  • The demo qualifies the transformation only; account verification, key creation, and a successful authenticated primary-value request are required for activation.
  • Inputs and outputs must be reviewed in the context of your own data rights, policies, and operational controls.
  • No ranking, accuracy, savings, recovery, prevention, customer, or revenue result is claimed.

From evaluation to first value

A short, explicit path

  1. Run the keyless demo.

    Confirm that the input and output shape match the job you need done.

  2. Create and verify a sandbox account.

    The email claim flow creates a product-scoped API key. No card is required for the free tier.

  3. Call the authenticated primary-value operation.

    After mandate issuance, a successful authenticated POST /v1/verify that returns a decision and receipt.

Create a sandbox key Read the API reference

Running this in production

Would you evaluate this for a production deployment at $299 per month?

The case this is built for. An agent acts for a principal who is not in the loop at the moment it acts. You need to know, at the moment of the action and provably afterwards, that the authority it claimed was the authority it was actually granted. A signed mandate makes that checkable by a party who did not issue it, which an ordinary approval rule in your own service cannot do.

If your agent only needs an approval gate you control, you do not need this. A policy check inside your own service is simpler, cheaper and sufficient. The signature is worth paying for when somebody other than you has to be able to verify the decision.

Published price, not a quote

  • Free — 500 verified actions per month, no card. Issuing mandates is free and never counts against it.
  • $299/month — 10,000 verified actions included, then $0.06 each, overage capped at $600 per period.
  • $1,499/month — 75,000 included, then $0.04 each.
  • $4,999/month — 400,000 included.

One unit is one action verified. These are the prices the checkout charges today. Full pricing.

You do not need to fill anything in to use this. The free tier, the docs and the walkthrough are open and need no conversation. This form exists only so that somebody who has already decided can say so, and so we can tell the difference between people reading the page and people who want to run it.





Unchecked by default. Leaving it unchecked still records your answer, and nobody will contact you.

Do not paste an API key, a mandate, a signature or any customer data into this form. None of it is needed and the fields will reject it.

We store the address and your two answers to judge whether this offer is worth running, and nothing else. No newsletter, no tracking pixel, no third party. See the privacy notice.