PR #153 went through ultracode review on the morning of 24 June. Five legal pages and /stamp-duty-calculator: the six surfaces the gsc-telemetry digest had been flagging as “URL is unknown to Google” across consecutive morning runs. The code review returned the same result on all six: HTTP 200, index, follow, canonical self-referencing, no broken prerender, no schema errors.
No technical blocker.
Phase 1.6 had opened with the standard working assumption: something in the infrastructure was wrong. The 20 June operator-side push addressed the obvious candidates. A sitemap canonical fix, twelve Request-Indexing submissions via GSC, IndexNow live. The gsc-telemetry skill I’d built that same day was parsing the Search Console coverage report into a morning digest. It kept confirming the pages were still unknown. The 24 June code review was the instrument for ruling out whatever remained.
It ruled out everything.
What the review checked
Five legal pages and /stamp-duty-calculator. HTTP response codes, meta directives, canonical tag implementation, prerender behaviour, schema validity. All passed.
The fix(seo) commit alongside the review captured genuine housekeeping I’d found: a prerender block cleared, the FAQ helper cleaned, the legal layout simplified. Real improvements to page quality. None of them explained why the six pages were unknown to Google. The audit established that before the housekeeping landed.
The earlier analysis of these six pages had established they had zero inbound links rather than thin content. That was a structural observation. The code review added what the structural observation couldn’t: a complete technical checklist. Not “I think the technical layer is clean” but “I’ve run the checklist and there is nothing here.” That distinction matters when you need to know which variable to change next. A hypothesis is a guess about cause. A clean audit run is a closed door.
The diagnosis
What I found isn’t a technical bug: footer-only pages.
/disclaimer and /stamp-duty-calculator appeared in the site footer. On every page. From the crawler’s perspective, those links existed. From the perspective of topical authority and crawl-budget prioritisation, they carried near-zero signal.
A footer link is navigational boilerplate. Googlebot recognises the pattern: present on every page, undifferentiated from dozens of others in the same template row. It follows footer links; it doesn’t assign them topical weight. A link in the body of an article on stamp duty or legal disclaimers is different. The surrounding text tells the crawler what the destination is about and why the author thought it warranted a link. That signal is what elevates a URL from “technically reachable” to “worth crawling.”
YMYL pages face additional E-E-A-T scrutiny. Legal disclaimers and financial calculators fall squarely into the category. A page that appears only in a footer, with no topical context linking to it, signals low editorial endorsement regardless of how correct the technical implementation is. The six pages weren’t thin. They weren’t blocked. They were stranded at the end of footer links that Googlebot had no particular reason to prioritise.
The fix
The 8-pillar content map had completed the day before, on 23 June. Six additional articles published in a single push, closing out the topic-authority backlog. The material was already live.
The chore(content) commit on 24 June: body-prose internal links from six existing articles into /disclaimer and /stamp-duty-calculator, at points in the text where those links were topically natural. Not new pages. Not structural changes. Changes to content that had already shipped, carrying links those articles should arguably have included on first publish.
This is crawl-weight transfer by the cheapest available mechanism. Footer links distribute PageRank across the entire site evenly. The destination gets a fraction of a fraction on every page load, with no topical signal attached. Body-prose links concentrate signal on a specific destination, with surrounding text that establishes relevance. Six articles. Two destination pages. One commit.
Why the sequence mattered
Technical audit first, editorial fix second. That order is deliberate.
Running the editorial fix before ultracode would have left open the possibility that the pages were unknown for a technical reason. If they remained unknown after the body-prose links landed, the diagnosis would be ambiguous: was the crawl-weight transfer insufficient, or was there still a technical blocker the editorial fix had obscured?
Ultracode first means the technical layer is certified clean before any editorial variable is introduced. If the pages index after the links land, the mechanism is clear. If they don’t, the next hypothesis can focus on crawl-budget allocation or content quality, not on whether the canonical tag was the problem all along.
The gsc-telemetry morning digest made the gap visible consistently. Without it, six unknown pages could have gone unnoticed for months. The default Search Console UI requires active investigation of specific URLs rather than surfacing coverage regressions as a recurring signal. The digest exposed the gap; the code review ruled out the technical layer; the editorial fix closed it. Three distinct instruments, one clean chain.
The twelve Request-Indexing submissions from 20 June are still processing. The 24 June commit is complete; the confirmation is pending.
What this changes in the workflow
Two process changes worth extracting.
First: the ultracode → editorial sequence should be the default for any page GSC flags as unknown despite a sitemap entry. The technical audit costs a PR review and a morning. Running it first removes the ambiguity that corrupts the diagnostic signal when an editorial fix doesn’t produce the expected outcome.
Second: body-prose link coverage belongs at page launch, not as a post-hoc fix. /disclaimer and /stamp-duty-calculator had been live before the content map filled out; the correct time to add those links was when the first topically relevant article shipped. The 24 June commit is catch-up. The process change: when a new page launches with only footer links, log it explicitly as a gap to close when the first relevant article ships. Don’t wait until GSC flags it.
Out of scope
Whether the six pages index after the fix is not established here. Confirmation is a GSC coverage report away, with a lag of days to weeks.
The AdSense “low value content” signal that surfaced on 23 June affects a different set of pages. Not related to the six unindexed pages. Out of scope.
