Start here
npx rigorrun demo replays a recorded run offline. Then test a Stripe refund
agent: a local twin with no keys, your agent over HTTP, a canary and a gate.
RigorRun documentation
Your agent acts for one customer at a time. RigorRun sends it tickets that tempt it across that line, then reads the system itself — first Stripe, your test mode or a local twin with no keys — and puts what actually happened beside what the agent said it did.
Connect your agent
A black-box agent: RigorRun posts each ticket to the endpoint your agent already serves and reads Stripe afterwards.
Understand the verdict
Verification strength is the thing to read first. A verdict
against a real system is PARTIAL, and that is honest rather than a defect.
Know the limits
Limitations is a real page with real entries. Three of the five evidence tiers are unreachable today, and it says so.
Not on Stripe? RigorRun can build a suite from one demonstration on an MCP server, an HTTP API or a web application.