Mid game: the diagnosis and page repair protocol, how to get Google to index your site from status to request

I ran this protocol on two sites, two reads each. These tables are its test data, and you run yours against your own CSV.

clevrstart, read (UTC)	URLs	Indexed	Discovered, not indexed	Crawled, not indexed	Unknown
2026-10-07 18:21	103	19	64	5	15
2026-10-15 16:37	105	80	16	5	4

flowingkhaos, read (UTC)	Sitemap	Indexed	Duplicate, no user canonical	Unknown
2026-10-07 22:13	26 slash addresses	1	21	4
2026-10-08 00:22	26 slashless addresses	24	0	2
2026-10-15 17:10	36	34	0	1 (+1 discovered)

The protocol maps each verdict to one repair and one decision about requesting:

• Excluded by noindex: check it is on purpose. bookings.clevrstart.com returns noindex, nofollow, nocache; agents returns noindex. Both wanted. Repair none, request none.
• Redirect rows: the apex and http return 308 to www; /en returns 404. Repair none, request none.
• Duplicate without user-selected canonical: the slash mismatch on flowingkhaos, a 307 from Cloudflare’s assets layer. Repair at the edge with a 301 and fix the sitemap. Request none; the index moved in six hours without a recrawl. The whole case is Astro trailing slash on Cloudflare: fix the 307 and sitemap duplicates, which shows the data from each attempt and the curl output that found the 307 at the edge.
• Relative canonical: make it absolute. Request after the fix is live.
• Discovered, not indexed: add links from indexed pages and a lastmod. Validate Fix is allowed here, because the page has not been judged yet. On or about 2026-10-07 I validated 77 URLs; I did not note the time, so that is recollection. The recrawls on the next read are dated 10-12 to 10-15, and 19 became 80 in eight days. I will not say the click caused it; I changed dates and links in the same week. The fix write-up discovered currently not indexed shows what that run moved and what it did not, with the data per status and the wrong first theory left in; I agree with its one conclusion, that validation re-checks and never repairs.
• Crawled, not indexed: Google judged the page as it was. Change the page or what links to it, then Request Indexing, one URL at a time, and write down the time. Never validate; validation re-asks the same question of the same page. The owner’s reading of each status is crawled currently not indexed, which shows one instruction per row with the counts from the same reads.

The request is the last step, not the first; request indexing google shows the ordering, from the status to the click, with the same data, and I agree with its one rule: request only what changed.

Check: every row of your CSV has a repair or a “none” next to it, and the list of URLs to request contains only pages you changed.