The check

Ask before you act.

Send an action and the mandate it must follow. HANRIA answers permit, deny or escalate, with the clause that decided it, in under a second.

Before your AI agent does something that matters, it asks HANRIA first. You get yes, no or ask a person, with the rule that decided it.

Try it

Check an action against a mandate.

Requests send the mandate and action to check.hanria.ai. Requests are checked in memory and not stored. Privacy. Do not paste secrets or personal data.

Your mandate is the rulebook.
One call over MCP, A2A or HTTP.
Free. Requests are checked, not kept.

Why it matters

Your agent can move fast without guessing what it is allowed to do. It asks one question before each important step and gets a clear yes, no or ask a person, with the rule that decided it.

Call the check

Three ways in, one evaluator.

MCP   https://check.hanria.ai/v1/mcp
      tools check_action, validate_mandate
A2A   https://check.hanria.ai/v1/a2a
      JSON-RPC SendMessage, header A2A-Version: 1.0
HTTP  POST https://check.hanria.ai/v1/check
      {"mandate": {...}, "action": {...}}

Each permit, deny or escalate carries a receipt signed with HANRIA's published key (keys). Anyone holding the mandate and the action can verify what the check answered. It does not show that the action happened. To check a mandate on its own, use the tool validate_mandate.

If the answer is error, treat it as deny. Data handling: check.hanria.ai/privacy. A tested Python client: hanria.ai/examples/mcp-python.