docs(evidence): close the duplicate specimen-CTA PR #107/#155 cluster on current state

.lane/reports/reconcile-specimen-cta-prs-107-155-2026-08-14.md

@@ -0,0 +1,57 @@
+# Lane report: reconcile/specimen-cta-prs-107-155-2026-08-14
+
+Date: 2026-08-14
+Lane: tinystudio-io lane 1
+Item: "[unreviewed-by-opus] Reconcile the duplicate specimen-CTA PRs
+#107/#155 — byte-identical specimen.html band hunks"
+
+## What I did
+
+Investigated the two-PR cluster against live GitHub state and found the
+premise had fully decayed — the fix has already landed and the cluster is
+already resolved:
+
+- **#107** (`fix/specimen-conversion-cta`): CLOSED 2026-08-14, never
+  merged. Its branch went conflict-locked against main; PR #155's
+  supersession note documents this.
+- **#155** (`fix/specimen-conversion-cta-lane1`): **MERGED into main**
+  2026-08-13 as commit `b81281f` ("fix(public): add in-content conversion
+  CTA to the /specimen proof page (#155)").
+- No open PR carries the fix: `gh pr list` shows no specimen-CTA delivery
+  path in the open set. This differs from the pricing-callout cluster
+  (#68/#114), where the fix had NOT yet landed and a surviving open PR
+  (#194) needed to be named.
+
+Verified the duplicate claim from the item: the two branches' hunks to
+`public/specimen.html` are byte-identical. `git diff` of each branch
+against its own merge-base (only the diff header/index lines differ, which
+is expected since the branches were cut from different main bases; the
+content hunks match exactly):
+
+- #107: `@@ -130,6 +130,13 @@` — adds the `.band` block (h2, copy,
+  `<a class="cta" href="/#start">Request the appraisal</a>`, note).
+- #155: `@@ -131,6 +131,13 @@` — the identical block, same 7 added lines
+  (one-line context shift because main moved between the two cuts).
+
+Confirmed the survivor is on main and deployed:
+
+- `git merge-base --is-ancestor b81281f origin/main` → true (origin/main
+  head `d981610`, 2026-08-14).
+- Live `https://tinystudio.io/specimen` serves the `.band`, the CTA
+  labelled "Request the appraisal", and the `/#start` target (verified by
+  fetch, 2026-08-14).
+- `npm run check` and `npm test` pass on origin/main (head `d981610`),
+  including the static source guard in `scripts/check-site.mjs` ("Specimen
+  in-content conversion CTA" block) that fails CI if the band, CTA, note,
+  or >=44px tap target regress.
+
+## Actions
+
+- No GitHub state change was needed: #155 is merged, #107 is already
+  closed, and no open duplicate delivery path survives.
+- Recorded the reconciliation closeout as evidence so the review queue
+  item can be closed as already resolved.
+
+## Evidence
+
+- docs/evidence/specimen-cta-prs-107-155-closeout-2026-08-14.md (this PR)

docs/evidence/specimen-cta-prs-107-155-closeout-2026-08-14.md

@@ -0,0 +1,105 @@
+# Duplicate specimen-CTA PRs #107/#155 — reconciliation closeout (2026-08-14)
+
+Date: 2026-08-14
+Scope: the review item "[unreviewed-by-opus] Reconcile the duplicate
+specimen-CTA PRs #107/#155 — byte-identical specimen.html band hunks".
+This receipt records a state verification of the repository's pull requests
+plus the reconciliation finding; it is process evidence, not a live-site
+measurement. The site-side verification of the fix itself is carried by the
+merged fix commit PR #155 and by the repo's standing `npm run check` and
+`npm test` suites, re-run on the current head.
+
+Prior receipts: `docs/evidence/specimen-in-content-cta-rereverify-2026-08-14.md`
+(PR #195, merged 2026-08-14) already re-verified the fix content on main
+and live; this receipt is the sibling review item's closeout for the PR
+cluster itself.
+
+## The item's premise (from the review queue)
+
+Two PRs carried the same fix — an in-content conversion CTA band on the
+/specimen proof page — with byte-identical `public/specimen.html` band
+hunks: PR #107 (`fix/specimen-conversion-cta`) and PR #155
+(`fix/specimen-conversion-cta-lane1`). A reviewer cannot tell which is
+canonical from the diff alone; the reconciliation must name the surviving
+delivery path and close the stale duplicate.
+
+## Current GitHub state (2026-08-14, this run)
+
+The premise has fully decayed: the fix has landed and no duplicate remains
+open.
+
+| PR | head branch | state (2026-08-14) | mergeability |
+|---|---|---|---|
+| #107 | `fix/specimen-conversion-cta` | CLOSED 2026-08-14T11:06:43Z, never merged | — |
+| #155 | `fix/specimen-conversion-cta-lane1` | **MERGED** 2026-08-13T22:09:21Z (`b81281f`) | merged |
+
+- **#155** is the surviving delivery path and already landed: its fix
+  commit `b81281f` ("fix(public): add in-content conversion CTA to the
+  /specimen proof page (#155)") is on `origin/main`
+  (`git merge-base --is-ancestor b81281f origin/main` → true; head
+  `d981610`, 2026-08-14).
+- **#107** was closed 2026-08-14 without merging. Its branch went
+  conflict-locked against main; PR #155's supersession note records that
+  #107 had carried the same change set since 2026-08-11 without landing
+  and was stale.
+- **No open PR carries the specimen-CTA fix.** The open-PR set
+  (`gh pr list --state open`, 2026-08-14) contains no specimen-CTA
+  delivery path.
+
+## The duplicate claim is confirmed: byte-identical specimen.html hunks
+
+The item says the two PRs carry byte-identical `public/specimen.html` band
+hunks. Verified by diffing each branch against its own merge-base with
+main, restricted to `public/specimen.html`:
+
+- PR #107 (`fix/specimen-conversion-cta`): hunk `@@ -130,6 +130,13 @@`,
+  +7 lines.
+- PR #155 (`fix/specimen-conversion-cta-lane1`): hunk `@@ -131,6 +131,13 @@`,
+  +7 lines.
+
+The two hunks are byte-identical in content: the `.band` block with the
+`<h2>That was a real one. Yours is read the same way.</h2>` headline, the
+four-passes copy line, `<a class="cta" href="/#start">Request the
+appraisal</a>`, and the no-guarantees `.note`. The only differences in the
+raw diffs are the git index lines and the one-line hunk context offset
+(#107 was cut from an earlier main; #155 from a later one) — the added
+lines match exactly. The companion files (`public/specimen.css` +7,
+`public/specimen.js` 1/1, `scripts/check-site.mjs` +29, and the locked
+heading-hierarchy outline) are also identical in shape across both PRs per
+their file lists.
+
+## The fix is on current main and live
+
+1. **Fix commit on current main** — `git merge-base --is-ancestor b81281f
+   origin/main` → true; origin/main head `d981610` (2026-08-14).
+2. **Source: the band CTA exists** — `public/specimen.html` lines 134-139
+   carry the `.band` block between the report and the footer with
+   `<a class="cta" href="/#start">Request the appraisal</a>` and the
+   no-guarantees note; the homepage routes its "Read the specimen"
+   call-out to `/specimen` (`public/index.html`).
+3. **Source: CI guard exists** — `scripts/check-site.mjs` ("Specimen
+   in-content conversion CTA" block) fails the build if the band, its CTA
+   to `/#start`, the no-guarantees note, `.band .cta` styling, or the
+   >=44px tap target (`padding:16px 24px`) regress.
+4. **Live site** — fetched `https://tinystudio.io/specimen` (2026-08-14):
+   the page serves `class="band"`, the "Request the appraisal" CTA, and
+   the `/#start` target (two occurrences each: nav CTA + band CTA, as
+   expected).
+5. **Test suite** — `npm run check` passes (exit 0) and `npm test` passes
+   on origin/main head `d981610`, including the specimen CTA guard.
+
+## Reconciliation finding
+
+**The review item is already resolved; no action was needed on GitHub
+state.** #155 merged the fix into main on 2026-08-13 and deployed; #107
+was closed on 2026-08-14 as the stale duplicate. The surviving delivery
+path is the merged PR #155 (commit `b81281f`). There is no open duplicate
+to close and no surviving open delivery path to name — the cluster is
+fully resolved, and the review-queue item can be closed as already
+resolved.
+
+## Evidence
+
+- This file: `docs/evidence/specimen-cta-prs-107-155-closeout-2026-08-14.md`
+- Lane report: `.lane/reports/reconcile-specimen-cta-prs-107-155-2026-08-14.md`
+- Related prior receipt (fix-content re-verify): `docs/evidence/specimen-in-content-cta-rereverify-2026-08-14.md`