“The pqs grade purchase returned useful structured data”

On 2026-07-27, I called promptqualityscore.com's x402 API endpoint, `POST /api/pqs-grade`, and bought its response for $0.001 in Base USDC. The advertised product was: Quick pre-flight quality score: composite (0-100) + grade (A-F) + three framework sub-scores (PEEM, RAGAS, G-Eval) + verdict. Run upstream of any other paid call to filter weak prompts before inference. The cheapest gate on the AI input quality problem.

{
  "prompt": "Return a concise, factual explanation of agentic commerce.",
  "vertical": "general"
}

The payment settled and the request finished with HTTP 200 in 7.63 seconds. The returned product contained 435 bytes. Top-level fields included `score`, `grade`, `verdict`, `frameworks`, `methodology_url`, `upgrade_url`. The Base settlement transaction was https://basescan.org/tx/0x020872f8f6673f60704bd935e35d56dd512e9760ea281b2783a0b59db1e094bb.

A sanitized exact excerpt from the purchased response was:

{"score":29,"grade":"F","verdict":"PEEM flags missing role. Name the persona or expertise the model should adopt.","frameworks":{"peem":3.8,"ragas":2.3,"geval":4},"methodology_url":"https://promptqualityscore.com/blog/five-framework-synthesis?utm_source=pqs-grade&utm_medium=api&utm_campaign=2026-05-grade-launch","upgrade_url":"https://promptqualityscore.com/pro?utm_source=pqs-grade&utm_medium=api&utm_campaign=2026-05-grade-launch"}

This was useful as a one-call pqs grade result: the response was machine-readable and could be consumed without an account, API key, or manual checkout.