agentleFS
geniushub-seogeniushub-seo/gsc-mcp/.agents/skills/index-health/SKILL.md

index-health

Skill7 starsChanged 3 months agoMITView on GitHub

Check whether Google has actually indexed a site's pages, and diagnose the ones it has not. Use when asked whether pages are indexed, why a page is not showing in search, about crawling problems, canonical issues, or sitemap errors.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Call list_sites to resolve the site_url, then:

    line 12
  2. Very specific:

    Always read truncated, scan_capped, and ordering on any query_search_analytics sample.

    line 84

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Index Health Check

A page that is not indexed earns nothing, no matter how good it is. This finds the ones Google is skipping and says why.

Step 1 — Property and sitemaps

Call list_sites to resolve the site_url, then:

manage_sitemaps
  site_url: <resolved>
  action: "list"

Report per sitemap: path, last_downloaded, is_pending, warnings, errors. A sitemap Google has not downloaded in weeks, or one carrying errors, explains a lot of missing pages before you inspect a single URL.

Step 2 — Pick which URLs to inspect

inspect_url accepts 1–10 URLs per call and is quota-limited per property per day. It is not a crawler — never point it at a whole site.

Choose deliberately:

  • User supplied a list — use it, up to 10 at a time.
  • User asked about the whole site — do not guess. Pull the pages that matter with query_search_analytics, dimensions: ["page"], sort_by: impressions, row_limit: 25, then inspect the top 10. Read truncated, scan_capped, and ordering before treating that list as complete. Explain that you sampled and why.
  • User asked about new pages — ask for the URLs. Pages with no impressions yet cannot be found through search analytics.

Step 3 — Inspect

inspect_url
  site_url: <resolved>
  urls: [<up to 10>]

The response carries quota_used_today and, near the limit, quota_warning. Surface the warning to the user rather than silently continuing.

Step 4 — Read the verdicts

For each URL the important fields are:

FieldWhat it tells you
verdictPASS / NEUTRAL / FAIL — the headline
coverage_stateThe specific reason, e.g. "Indexed", "Crawled - currently not indexed", "Discovered - currently not indexed"
indexing_stateWhether indexing is blocked, e.g. by a noindex
robots_txt_stateWhether robots.txt allowed the crawl
page_fetch_stateWhether Google could fetch the page at all
google_canonical vs user_canonicalWhen these differ, Google ignored your canonical — a common and invisible cause of missing pages
last_crawl_timeStale means Google has not looked recently
sitemapWhich sitemaps reference this URL; empty means it is not in any

The optional blocks (mobile_usability_verdict, rich_results_verdict, amp_verdict) are omitted by Google when they do not apply. Their absence is not a failure.

Step 5 — Report as an action list

Group by what the user has to do, not by URL:

## Not indexed — needs action
- <url> — Crawled but not indexed. Google fetched it and chose not to index.
  Usually thin or duplicate content. Compare against <the canonical Google chose>.

## Canonical mismatch
- <url> — You declared <user_canonical>, Google chose <google_canonical>.
  Your declaration is being ignored; check for duplicate content or conflicting signals.

## Blocked
- <url> — robots.txt disallows it / page carries noindex.

## Not in any sitemap
- <url> — Add it so Google discovers it faster.

## Indexed and healthy
- <count> URLs, no action needed. (List them only if asked.)

Accuracy rules

  • Always read truncated, scan_capped, and ordering on any query_search_analytics sample. truncated=true means more pages exist; scan_capped=true means even the scan was incomplete.
  • This reports what Google has stored, not the live page. Fix a page today and this still shows the old verdict until Google recrawls. Say so whenever the data looks stale.
  • It is not a live test, does not request indexing, and does not enumerate every indexed URL. It cannot replace the Page Indexing report in the Search Console UI.
  • Quotas are per property per day; see https://developers.google.com/webmaster-tools/limits. Do not state a specific number — Google changes them.
  • "Discovered - currently not indexed" and "Crawled - currently not indexed" mean different things. The first means Google knows the URL but has not fetched it (often a crawl-budget or internal-linking issue); the second means it fetched and declined (usually a content issue). Do not give the same advice for both.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.