x402 Seller: Payable and Listed: test results
Tested 2026-09-30, skill version 1.0.0 at the time of loading this page. We run every skill on a strong and a weak model before it is listed, and publish both verdicts, including where the weak one fails.
Verdicts
- Date
- 2026-09-30
- Strong · Claude Sonnet (claude-sonnet-5-5, Claude Code alias "sonnet")
- Named the planted defect in all 6 cases in every run (missing PAYMENT-REQUIRED header, missing extra, extensions placed inside accepts[0], test network plus x402.org facilitator, UTF-8 versus atob decoding, the 30-day delisting), also found a second defect where present, wrote the German case in German, and never promised a listing. Sometimes adds cautious extra remarks (a placeholder address, the missing bazaar block on a body that never claimed one).
- Weak · Claude Haiku (claude-haiku-4-5-20251001, Claude Code alias "haiku")
- Named the planted defect in 6 of 6 cases on the second run. On the first run it missed the missing extra field entirely and reported only a masked placeholder wallet address as the defect; a rule to tick every field and treat placeholders as redaction fixed that in the next run. Still breaks the no-code-block rule (shows a small JSON snippet), quotes the skill file back to the reader, and once used up to 9,000 tokens on one case. Weaker on the German case wording (verdict line not in the requested form). Slightly over-promises timing of a listing (says it should appear within minutes to hours).
Note
6 planted-defect cases (one in German), 2 full runs per model plus one single-case re-run after a wording change; mechanical checks pass 12 of 12 on the final full run, 11 of 12 on the first. The skill reads pasted text only and cannot confirm live headers, settlement history or edge settings. Facts come from our own measurements on Base mainnet plus public issue threads; listing behaviour of the CDP catalog can change and is never guaranteed.
What was not measured
- Models other than the two named above were not run.
- Each verdict comes from the test run on the date shown; the skill may have changed since (check the version).
- Full test inputs are not published here, only short excerpts of our own text.
- Results on your own texts, languages and domains can differ.