a wontfix reason-derived grounds value skips ValidateText so a control character in wontfix_reason still reaches yamlScalar

a wontfix reason-derived grounds value skips ValidateText so a control character in wontfix_reason still reaches yamlScalar

Surfaced by the itd-179 round-3 builder as an observation and captured rather than fixed, because it is pre-existing and outside the round's records.

A wontfix whose grounds are DERIVED from the reason bypasses grounds.ValidateText, so the control-character refusal added in c452ee55 does not cover that path: a control character in a wontfix_reason still reaches yamlScalar.

Why it is a nitpick rather than a defect: that path mints nothing, so it fails atomically with no orphan draft — which is the consequence the sibling record (iss-2608301206032013) existed to prevent. The refusal happens either way; only its site and its message differ.

Left open for the facilitator. Closing it would mean routing derived grounds through the same validator as supplied grounds, which is a small consolidation in the direction the one-canonical-primitive rule already points.

Grounds

• pursued: a wontfix reason carrying a control character is refused as a grounds refusal naming the rune; a refusal arriving from the frontmatter serialiser instead would show it wrong