Assess

On-demand paid AI assessment of a service shortlist

Runs a FRESH flagship-LLM comparison of an already-assessed shortlist of services for a stated need, and charges x402 only for that fresh reasoning. It never charges to read an already-computed assessment (that stays free on the service detail). The report is a modular envelope: data carries one block per priced module (today the advisor answer under data.answer), and meta carries report_version and the modules_run list, so future modules are additive and never breaking. The paid flow is the standard x402 handshake: the first call (no PAYMENT-SIGNATURE) returns HTTP 402 with a PAYMENT-REQUIRED header and an accepts[] body (single option, $0.25 USDC on Base); sign the payment and retry the same request with a PAYMENT-SIGNATURE header. On success the 200 carries a PAYMENT-RESPONSE header. You may optionally include a probe target (probe { slug, endpoint_path? }) to also test one listed service live: after the fresh reasoning the server makes a real x402 payment to that endpoint and analyzes the response, and the single accepts[0] amount is then $0.25 plus that endpoint price X. The report then carries a probe_report block with a verdict plus truncated extracts (never the verbatim third-party body). The probe is only offered when live probing is armed, and probe fees are non-refundable regardless of outcome. If the fresh run cannot be produced (an unresolvable shortlist or a model miss) the endpoint answers 503 BEFORE settling, so the caller is never charged. When the payment layer or the on-demand assessment feature is not configured, the endpoint answers 503. There is NO refund. This endpoint is agent-first (JSON only).

post/assess

Request

Base URL: https://x402-list.com/api/v1
URL: https://x402-list.com/api/v1/assess
Auth: none declared

Request body

questionstring required

The need to assess the shortlist against (1 to 1000 characters).

servicesstring[] required

Service slugs to compare (1 to 8, de-duplicated, order kept).

Response

The modular assessment report. data.answer is the AI-marked recommendation; meta carries report_version and modules_run. The PAYMENT-RESPONSE header carries the settlement receipt.

Example response

{
  "provenance": {
    "attribution": "Data: x402-list.com (CC BY 4.0)",
    "cite_as": "https://x402-list.com/services/acme-generate",
    "source": "https://x402-list.com"
  }
}

Changes

No recorded changes to this operation across all 1 revision of this API.