Home › Use cases — Liceat

What you can do with agent warrants

The everyday jobs of running autonomous agents safely — each solved with a scoped, provable warrant:

Set agent spend limits

Cap how much an AI agent can spend with a stateful, cumulative spend limit enforced across every cal

Revoke agent access

Instantly revoke an AI agent’s authority by warrant id — and have it cascade to every sub-agent the

Audit agent actions

Get a tamper-evident, verifiable record of every action an AI agent took — what it did, under whose

Delegate to sub-agents

Let one AI agent delegate authority to another safely: a delegated warrant can only narrow, never es

Rate-limit agents

Bound how often an AI agent can act with a sliding-window rate limit enforced per warrant — stop run

API keys for agents

API keys grant all-or-nothing access and can’t be scoped per task. Warrants give AI agents narrow, s

Prevent runaway spend

Stop an AI agent from spending beyond intent with a cumulative spend cap, rate limit, and expiry enf

Human approval gates

Require human approval for high-stakes AI agent actions by issuing warrants that only permit low-ris

Scope agent permissions

Grant an AI agent exactly the permissions it needs — specific resources and actions, nothing more —

Agent action logging

Produce a complete, verifiable, exportable record of AI agent actions for SOC 2, the EU AI Act, and

Start in 60 seconds.

Issue a scoped, revocable warrant and enforce it at the point of action.

Get an API key Docs