fix(seo): cut retired Agent Desk title consolidation wires

docs/evidence/agent-desk-retired-title-2026-08-11.md

@@ -0,0 +1,106 @@
+# Retired Agent Desk title/snippet: cut the canonical consolidation wire
+
+Date: 2026-08-11
+Scope: `public/agent-desk.html` (head canonical/og:url), `src/worker.js`
+(retired 410 response copy), `scripts/check-site.mjs` (regression guard).
+Follow-up to `docs/evidence/agent-desk-retired-title-2026-08-09.md`, which
+added `noindex, nofollow` and retired framing to the legacy surface. This
+receipt records the second, deterministic pass: the two remaining wires that
+could still feed the retired "TinyStudio Agent Desk" title/snippet to Google
+were cut. It is not a live Google-index measurement and claims nothing about
+when or whether Google's search results change.
+
+## Why this pass exists
+
+Google still presents the retired self-serve "TinyStudio Agent Desk"
+title/snippet for tinystudio.io. The 2026-08-09 pass is verified live (both
+fetches below), so the stale presentation can no longer come from the legacy
+page's own title or description — those are gone. Two index-facing wires
+remained:
+
+1. **The canonical/og:url wire.** The 2026-08-09 receipt itself diagnosed the
+   original mechanism: "the legacy page's title consolidated through its
+   canonical" onto the homepage URL. The pass added `noindex, nofollow` but
+   left `<link rel="canonical" href="https://tinystudio.io/">` and
+   `<meta property="og:url" content="https://tinystudio.io/">` in place.
+   Live-before state (2026-08-11):
+   `curl -s https://tinystudio.io/agent-desk` served `noindex, nofollow`,
+   the retired title/description — and still declared the homepage as the
+   page's canonical and og:url. A noindex page whose canonical names a live
+   page tells Google it is a duplicate of that page: exactly the consolidation
+   that handed the retired title to tinystudio.io, still declared on every
+   recrawl until the removal is processed. Google's own guidance is that a
+   page you do not want indexed should not canonicalize to another URL.
+2. **The retired-410 copy wire.** `app.tinystudio.io` and `api.tinystudio.io`
+   respond 410 Gone with the copy "TinyStudio.io now runs the self-serve
+   Agent Desk from the main domain." Live-before state (2026-08-11): both
+   fetches returned that sentence. It re-presents the retired self-serve
+   product as the current thing the main domain runs, and the pages stay
+   crawlable until the 410 is processed.
+
+## What changed
+
+1. `public/agent-desk.html` head: the canonical and og:url now name the
+   legacy page itself (`https://tinystudio.io/agent-desk.html`) instead of the
+   homepage. The consolidation wire is cut: no indexed page points the legacy
+   surface at `tinystudio.io/` any more, so a recrawl can no longer merge the
+   retired page's signals or title onto the homepage URL.
+2. `src/worker.js`: the retired 410 responses (app and api) no longer claim
+   the main domain runs the self-serve Agent Desk. They now say the app and
+   the self-serve Agent Desk are both retired and that TinyStudio.io's
+   current offer is the Website Appraisal.
+3. `scripts/check-site.mjs`: the retired-desk dogfood guard now fails `npm
+   run check` when the legacy page's canonical or og:url points anywhere but
+   the legacy page itself (or is absent), and when the worker's retired 410
+   copy claims the main domain "now runs the self-serve Agent Desk" — so
+   neither wire can silently return.
+
+## What deliberately did not change
+
+- The legacy surface itself: `agent-desk.html` remains a working self-serve
+  page served at `/agent-desk` and `/agent-desk.html`, still `noindex,
+  nofollow`, still retired-framed. Old deep links and bookmarks keep working.
+- `robots.txt`, `sitemap.xml`, the homepage, and the other public pages:
+  their titles and descriptions were already current.
+- No redirect was added for `/agent-desk`: the page stays reachable, it is
+  just no longer indexable or canonicalizable onto the homepage.
+
+## What is tested and what is not
+
+Tested: the extended guard fails loudly on regression (the canonical and the
+410 copy are asserted directly), and the full suite passes on this commit —
+`npm run check` green, `npm test` green, exit code 0.
+
+Not tested: whether or when Google drops the old title/snippet from its
+results. De-indexing and re-consolidation are live processes that start only
+after Google recrawls the pages; nothing here implies a ranking, visibility,
+lead or revenue outcome. The honest measure is a future controlled re-run
+recorded through `evidence-fixtures/ai-search` (or Search Console), once
+enough time has passed for a recrawl.
+
+## Verification (reproduce)
+
+```
+npm run check
+npm test
+git diff --check
+```
+
+All three checks pass on this commit. Live-before state (2026-08-11):
+
+```
+curl -s https://tinystudio.io/agent-desk      # noindex + retired head, canonical/og:url still -> homepage
+curl -s https://app.tinystudio.io/            # 410 body still claimed the main domain runs the Agent Desk
+curl -s https://api.tinystudio.io/            # 410 JSON message made the same claim
+```
+
+After deploy, the same fetches should show the self-canonical
+`https://tinystudio.io/agent-desk.html` and the corrected 410 copy.
+
+## Closeout
+
+This closes the repository-side item against current main: the legacy Agent
+Desk surface no longer declares itself a duplicate of the homepage through
+its canonical or og:url, the retired 410 responses no longer present the
+retired self-serve Agent Desk as the main domain's current offer, and CI
+fails if either returns.

public/agent-desk.html

@@ -15,8 +15,8 @@
       content="The self-serve Agent Desk is retired and is not the current offer. TinyStudio's current offer is the free leak audit of high-ticket service homepages."
     />
     <meta property="og:type" content="website" />
-    <meta property="og:url" content="https://tinystudio.io/" />
-    <link rel="canonical" href="https://tinystudio.io/" />
+    <meta property="og:url" content="https://tinystudio.io/agent-desk.html" />
+    <link rel="canonical" href="https://tinystudio.io/agent-desk.html" />
     <meta property="og:image" content="https://tinystudio.io/og-image.png" />
     <meta property="og:image:width" content="1200" />
     <meta property="og:image:height" content="630" />

scripts/check-site.mjs

@@ -1076,6 +1076,41 @@ if (!/\bretired\b/i.test(retiredDeskDescription)) {
   failures.push("Retired Agent Desk description must frame the surface as retired.");
 }
 
+// The legacy page must never canonicalize to the homepage or any other live
+// page: the canonical-to-homepage was the consolidation path that handed the
+// retired page's title to tinystudio.io in the first place (see the receipt
+// above), and a noindex page pointing its canonical at a live page tells
+// Google it is a duplicate of that page while the noindex removal is
+// processing. The canonical must name the legacy page itself (or be absent),
+// and og:url must not name any other page either.
+const legacyPageUrl = "https://tinystudio.io/agent-desk.html";
+const retiredDeskCanonical =
+  retiredDeskHead.match(/<link\b[^>]*\brel\s*=\s*["']canonical["'][^>]*>/i)?.[0] ?? "";
+const retiredDeskCanonicalHref =
+  retiredDeskCanonical.match(/\bhref\s*=\s*["']([^"']*)["']/i)?.[1] ?? "";
+if (retiredDeskCanonicalHref && retiredDeskCanonicalHref !== legacyPageUrl) {
+  failures.push(
+    `Retired Agent Desk canonical must be self-referencing (${legacyPageUrl}) or absent; found ${retiredDeskCanonicalHref}.`
+  );
+}
+const retiredDeskOgUrl =
+  retiredDeskHead.match(/<meta\b[^>]*\bproperty\s*=\s*["']og:url["'][^>]*>/i)?.[0] ?? "";
+const retiredDeskOgUrlHref = retiredDeskOgUrl.match(/\bcontent\s*=\s*["']([^"']*)["']/i)?.[1] ?? "";
+if (retiredDeskOgUrlHref && retiredDeskOgUrlHref !== legacyPageUrl) {
+  failures.push(
+    `Retired Agent Desk og:url must be self-referencing (${legacyPageUrl}) or absent; found ${retiredDeskOgUrlHref}.`
+  );
+}
+
+// The worker's retired 410 responses (app.tinystudio.io / api.tinystudio.io)
+// must not re-present the retired self-serve Agent Desk as what the main
+// domain "now runs": that phrasing claims the retired product is the current
+// offer of tinystudio.io and stays crawlable on 410 pages until they drop out
+// of the index.
+if (/now runs the self-serve Agent Desk/i.test(worker)) {
+  failures.push("Worker retired 410 responses must not claim the main domain now runs the self-serve Agent Desk.");
+}
+
 // ---- Meta descriptions (dogfood) -------------------------------------------
 // The leak audit this site sells flags a homepage whose served HTML carries no
 // description, so the site's own five public pages must not carry the same

src/worker.js

@@ -1266,7 +1266,7 @@ function retiredAppResponse() {
   <body>
     <main>
       <h1>TinyStudio app retired.</h1>
-      <p>The old TinyStudio app has been retired. TinyStudio.io now runs the self-serve Agent Desk from the main domain.</p>
+      <p>The old TinyStudio app has been retired, and so has the self-serve Agent Desk. TinyStudio.io's current offer is the Website Appraisal: the free leak audit of high-ticket service homepages.</p>
       <a href="https://tinystudio.io/">Go to TinyStudio.io</a>
     </main>
   </body>
@@ -1288,7 +1288,7 @@ function retiredApiResponse() {
       {
         ok: false,
         status: "retired",
-        message: "The old TinyStudio API has been retired. TinyStudio.io now runs the self-serve Agent Desk from the main domain.",
+        message: "The old TinyStudio API has been retired, and so has the self-serve Agent Desk. TinyStudio.io's current offer is the Website Appraisal: the free leak audit of high-ticket service homepages.",
         publicSite: "https://tinystudio.io/"
       },
       {