Bound how often an agent can act with a sliding-window rate limit carried in the warrant itself.
Infrastructure rate limits are per-IP or per-key. Agent risk is per-authority: this specific delegated task should make at most N calls a minute, regardless of where it runs. A rate_limit caveat encodes that in the warrant; the gateway enforces it as a sliding window against real call history.
{ type: 'rate_limit', max: 10, per: 60 } // ≤ 10 allowed actions per 60sA buggy or adversarial agent that enters a tight loop is throttled at the authority boundary — combined with a spend cap and an expiry, the blast radius of "the agent went haywire" is bounded in advance.
Liceat is permission infrastructure for AI agents. You issue a warrant — a scoped, spend-capped, revocable token — that grants an agent narrow authority; the agent presents it at the point of action; Liceat enforces it (scope, spend, rate, holder-binding, revocation) and records every decision in a tamper-evident log. It is open source with a zero-dependency SDK for TypeScript and Python.
The SDK and self-hosted gateway are free and open source. The hosted gateway at liceat.com gives 10,000 free decisions with no card, then meters at fractions of a cent per decision — payable by card or USDC.
Free to self-host. 10,000 free decisions on the hosted gateway — no card.
Get an API key Read the docs