Sprint C4: vendor-price verification card + eight ideas relabelled blocked-on-arrival

[pftg]

Follow-up to #508. Docs-only. Answers "what else is left from the roadmap to improve the course" by scheduling the one item worth scheduling and correcting the status of eight that were mislabelled.

I made this call rather than waiting on an answer — both changes are docs-only and reversible. Close this PR if you disagree.

Sprint C4 — vendor-price verification

The only remaining roadmap item that degrades on its own. 28 course files carry $X/mo-shaped figures; 55 name a vendor (Lovable, Supabase, Mixo, Carrd, Bubble, Stripe). 40.04 already flagged it — "Lovable price drift survived two previous sweeps" — and asked for an annual refresh cadence that was never added to any tracker.

The card leads with the distinction that makes it safe to hand off. The two kinds of dollar figure are mixed inside single files:

Kind	Example	Action
Vendor price	validation-tools-field-guide L86 — "VenturusAI … Pro from $10/mo"	Verify against live pricing page
Narrative / teaching	same file L111 — "Subscription model at $200/mo" (Mia's fictional clinic app)	Do not touch

A naive "refresh the prices" pass would corrupt teaching examples and break walkthrough consistency wherever a narrative number recurs across chapters.

Two of the 28 are SVGs, not markdown — mom-test-script.svg and good-vs-bad-answers.svg. Prices inside artwork are invisible to every text ratchet we run. That is the third instance of the exhibit-blindness trap in a month (the paid-pilot bar chart yesterday, the LI asset before it), so the card carries the re-export requirement explicitly.

Deliverable is a canonical table with per-row source URL and checked-date, not just edits — that was 40.04's own conclusion, and it makes the next refresh a diff rather than a re-derivation.

Eight ideas relabelled: "deferred" → BLOCKED ON ARRIVAL

Ch 2.2 sidebar · Ch 2.4 resistance · 1.2a builder-path split · template timing guide · deferred briefs · Option 2 bridge · post-ship retro · pilot testimonials.

Every one is gated on reader/GA4 data. The 28-day funnel is start_course 5, checkpoint-reveal 4, branch-click 2, glossary 1. You cannot derive a Carrd-fallback rate, a finisher cohort, or "demonstrated GA4 demand" from five course starts.

"Deferred" reads as ready when we get to it, and they have been carried that way across several sessions as if a clock were running. They are waiting on distribution — the same constraint the LinkedIn cards test. The note explicitly forbids the tempting workaround: shipping a structural change on 5 sessions of data is guessing with extra steps, and the original deferrals were right to demand evidence.

Net effect on the course roadmap

• Scheduled: C4 (this).
• Opportunistic: 31 SVG floor candidates, render-gate per image, never batch.
• Blocked on arrival: the eight above.
• Closed: course SEO/AEO (Course: palette pass, discovery diagnosis that retracts the plan's premise, GA4 setup closed #508), certificate, video, mail list.

Gates: docs-only, bin/hugo-build green.

🤖 Generated with Claude Code

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: e9d4f7eb-5803-4eb7-806b-59df37a44eae

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}