Skip to content

feat(copy): tokenize hero CTA + proof-card titles as enforced micro-copy#85

Merged
bdelanghe merged 1 commit into
mainfrom
feat/tokenize-microcopy
Jun 29, 2026
Merged

feat(copy): tokenize hero CTA + proof-card titles as enforced micro-copy#85
bdelanghe merged 1 commit into
mainfrom
feat/tokenize-microcopy

Conversation

@bdelanghe

Copy link
Copy Markdown
Contributor

What

First slice of prx-gwr8 — promotes four micro-copy strings to canonical content/strings.json tokens, each asserted by a @marketing scenario in the new content/marketing.feature.

content.mjs --check now enforces both directions: the page can't drift from the token, and the token must appear on the page (copy coverage 7/7). Copy stays readable inline — the token is the source of truth. This is the project's existing Gherkin-coverage model (same mechanism as the brand tagline/description), so no templating or placeholders.

Tokens: cta-read-bet, proof-{guest-room,prx,claude-box}-title.

Scope / rationale (your tokens question)

  • Micro-copy → tokens, yes (this PR). Section eyebrows + remaining labels follow the same recipe.
  • Long body prose → not tokenized. Inline markup (<strong>/<em>/<a>) makes flat tokens the wrong tool; it stays audited via the string-audit catalog.

Verification (native nix)

content.mjs --check (token + coverage gate), build, structure-audit, semantic, SHACL, reader, seo — all green locally.

🤖 Generated with Claude Code

First slice of prx-gwr8. Promotes four micro-copy strings to canonical
content/strings.json tokens, each asserted by a @marketing scenario in the
new content/marketing.feature. content.mjs --check now enforces both
directions — the page can't drift from the token, and the token must appear
on the page (copy coverage 7/7). Copy stays readable inline; the token is the
source of truth (no templating, no placeholders — the project's Gherkin-
coverage model, same as the brand tagline/description).

Tokens: cta-read-bet, proof-{guest-room,prx,claude-box}-title.

Long body prose (hero lead, the bet, honesty lead) intentionally NOT tokenized
— inline markup makes flat tokens the wrong tool; it stays audited via the
string-audit catalog. Remaining micro-copy (section eyebrows, existing
untokenized labels) follows this same recipe.

content/build/structure/semantic/shacl/reader/seo verified green locally.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@bdelanghe bdelanghe merged commit fcb1d37 into main Jun 29, 2026
10 checks passed
@bdelanghe bdelanghe deleted the feat/tokenize-microcopy branch June 29, 2026 01:08
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant