Build LogContent Pipeline
#76
2026-07-305 min read

The day-three mortgage draft cleared verification. The cluster had already shipped on day two.

On day three of a fourteen-day content run, the cron produced a mortgage rate-spike draft that cleared sentence-level verification. The same cluster had shipped the morning before. Sentence verification and cluster novelty are different problems — one belongs at the verification layer, the other doesn't.

Read post
Build LogConversion
#77
2026-07-305 min read

The thirteen-session visitor who reshaped three pages

GA4 turned up essentially one Upwork-referral visitor with more than thirteen returning sessions. Seven minutes on /work/, no action taken. The sample size was one. The behaviour was deliberate enough to act on anyway. Three site changes followed in a single sprint.

Read post
Build LogContent Pipeline
#78
2026-07-295 min read

Mortgage scored 20 and nothing shipped: the cron's zero-drafts firing is the verifier-corrective gate working

On Day 4 of 14, the cron ranked 150 clusters, placed mortgage first at calculator score 20, fetched the candidates, and produced zero drafts. That was correct. The verifier-corrective gate suppresses the draft when the ground is stale — silence is cheaper than shipping a duplicate pillar and retracting it.

Read post
SEOBuild Log
#79
2026-07-295 min read

Telemetry surfaced six invisible YMYL pages. The operator-side fixes didn't move them. Internal linking did.

After ga-telemetry and gsc-telemetry made six stranded YMYL pages visible, Phase 1.6's first move was operator-side: sitemap canonical fixes, IndexNow, twelve Request-Indexing nudges. The technical audit cleared every blocker. The pages stayed unknown. Internal linking — crawl-weight transfer from body prose — was the diagnosis the data forced.

Read post
Build LogSEO
#80
2026-07-294 min read

The og:image that wasn't there: blank WhatsApp previews and a pre-ship check no CI runs

Sharing a captainrandom.co.uk album link on WhatsApp returned a blank white card. The site had structured data, a navigation spec authored by 16 parallel agents, and data-driven course stubs. It had no default og:image. The gap was invisible until a real phone surfaced it.

Read post