01

The Lighthouse Score for Agentic Commerce.

Test every branch of your ACP and UCP checkout flows

Prova simulates AI shopping agents against your endpoints—surfacing broken flows, stale inventory, and protocol violations so you can fix them before they cost you revenue.

Agent Readiness Score

82 / 100 GOOD
Flow 92
Compliance 78
Performance 85
Errors 70
Data 83
3 critical issues · 7 warnings 73rd percentile

Supports ACP (OpenAI/Stripe) · UCP (Google/Shopify) · MCP

AI-referred traffic to retail sites grew 1,200% in 2025.
02

AI agents don't give second chances.

When a human hits a broken checkout, they retry. When an AI agent hits one, it silently moves on. The transaction is lost, and the agent never comes back.

1,200%

growth in AI-referred traffic to retail sites in 2025

805%

surge in AI traffic on Black Friday 2025 alone

1 in 5

Cyber Week 2025 orders involved an AI agent

86%

lower conversion from agent traffic vs. affiliate traffic

The highest-profile failure proves it.

OpenAI launched Instant Checkout in ChatGPT in February 2026. By March, they killed it. Only 12 merchants went live. Near-zero completed purchases. The root causes?

1

Stale inventory — agents showed products that were out of stock, producing silent cart errors.

2

Missing sales tax — a compliance gap across US states that any systematic test would have caught.

3

Fraud false positives — fraud tools couldn't tell legitimate agents from bots, blocking real purchases.

4

Session escalation failures — when checkout required human input, the handoff broke silently.

5

No idempotency — retry logic caused duplicate sessions that silently reduced conversion.

Every one of these failures was detectable by testing before production.

"Escalation mishandling is a larger revenue risk than payment decline rates."

— Checkout.com, Agentic Commerce Analysis, 2026
03

Three steps. Two minutes. Every flow branch tested.

1

Connect Your Endpoint

Paste your agent's URL or upload an OpenAPI spec. Prova auto-discovers every product lookup, cart, and checkout flow your agent exposes.

2

Prova Simulates Real Shoppers

Synthetic personas browse, search, add to cart, and check out — hitting every branch, edge case, and error path your agent can encounter.

3

Get Your Agent Readiness Score

A single 0–100 score with a detailed breakdown across six categories so you know exactly what to fix before going live.

Category What We Test
Flow Completeness Every browse → cart → checkout path resolves without dead ends.
Protocol Compliance Responses conform to MCP, OpenAPI, or your custom schema contract.
Data Freshness Prices, inventory counts, and product metadata stay in sync with source.
Error Handling Graceful recovery from bad input, timeouts, and upstream failures.
Performance P50 / P95 latency and throughput under simulated concurrent load.
Edge Cases Out-of-stock items, currency mismatches, oversized carts, and more.

Alpine Gear Co.

NEEDS WORK
67 / 100
Flow Completeness 75
Protocol Compliance 82
Performance 91
Error Handling 45
Data Quality 62

3 critical issues · 4 warnings

Sample Audit

See what a real audit looks like.

We ran Prova against Alpine Gear Co.'s ACP endpoint. Found 3 critical issues including a 500 error on out-of-stock items, ignored idempotency keys, and silent address validation failures.

Every one of these would have cost real revenue in production.

04

Why Prova?

01

Protocol-Specific Testing

Not generic QA. We test ACP and UCP checkout flows against the actual protocol specifications.

02

Synthetic AI Agent Personas

Not scripted bots. Real LLMs (GPT-4o, Claude, Gemini) with randomized shopping personas stress-test your endpoints.

03

Lighthouse-Style Scoring

A single Agent Readiness Score (0-100) with category breakdowns. Benchmark against other stores.

04

CI/CD Integration

Run Prova on every deploy. Get Slack alerts when your score drops. Coming soon on Pro plans.

Frequently Asked Questions

What is an Agent Readiness Score?
An Agent Readiness Score is a composite metric from 0 to 100 that evaluates how well your commerce endpoint handles AI agent interactions. It grades across five categories—Discovery, Cart & Checkout, Payment Handling, Fulfillment, and Error Recovery—giving you a single, actionable number similar to a Lighthouse score for web performance.
Do I need to change my code?
No. Prova tests your endpoint entirely from the outside, just like a real AI agent would. There’s no SDK to install, no code to embed, and no instrumentation required. Point us at your storefront or API, and we handle the rest.
Which protocols do you support?
Prova currently supports three agentic commerce protocols: ACP (Agent Commerce Protocol) for streamlined checkout flows, UCP (Unified Commerce Protocol) for full lifecycle management, and MCP (Model Context Protocol) for tool-use integrations. We add support for new protocols as they emerge.
How is this different from Postman?
Postman tests deterministic API contracts—you send a request, you expect a specific response. Prova simulates the non-deterministic behavior of real AI agents, including ambiguous queries, mid-session corrections, partial cart abandonment, and unexpected navigation paths that traditional API testing tools never exercise.
Will Prova make real purchases?
No. All tests run in a sandboxed environment. We never submit real payment credentials, complete actual transactions, or interact with live payment processors. Every checkout flow is simulated end-to-end without any financial risk.
How many API calls per test?
A typical test run makes between 15 and 30 API calls, roughly equivalent to one human browsing session. This includes product discovery, cart manipulation, checkout initiation, and error recovery scenarios. The exact count depends on which test suite you run.
What's the difference between ACP and UCP?
ACP (Agent Commerce Protocol) focuses specifically on the checkout experience—product selection through payment completion. UCP (Unified Commerce Protocol) covers the full commerce lifecycle, including post-purchase flows like order tracking, returns, refunds, and subscription management.
Why did OpenAI kill Instant Checkout?
OpenAI’s Instant Checkout launched with only 12 merchants, suffered from stale product data, couldn’t reliably prevent fraud at scale, and lacked graceful escalation paths when the agent couldn’t complete a task. These are exactly the classes of failures Prova is designed to detect before they reach production.
Can I run in CI/CD?
Yes. On the Pro plan, you can integrate Prova directly into your CI/CD pipeline. We provide first-class support for GitHub Actions and GitLab CI, with a CLI that returns structured exit codes so you can gate deployments on your Agent Readiness Score.
How do you simulate different AI agents?
We run tests using multiple frontier models—including GPT-4o, Claude, and Gemini—each configured with randomized personas, varying levels of domain knowledge, and different interaction styles. This ensures your endpoint works reliably regardless of which AI agent your customers use.
Is my endpoint data secure?
Yes. API keys and credentials are held in memory only for the duration of a test run, encrypted at rest using AES-256, and subject to a 90-day automatic deletion policy. We never store response payloads beyond what’s needed to generate your report.
?

Your agents are already shopping. Is your checkout ready?

Agent-driven commerce traffic is projected to grow 1,200% over the next two years. Businesses that aren't ready will lose sales to those that are. Prova gives your checkout the machine-readable layer it needs so AI agents can discover, validate, and complete purchases — without friction.

MCP & A2A protocols supported Sandbox-only environment SOC 2 compliance planned