What rules does the Muginai site audit cover?

Muginai audits 40+ on-page rules across six categories: structured data (schema.org validation), hreflang consistency, Core Web Vitals (LCP, CLS, FID), internal linking (orphan pages, anchor diversity), redirect chains (3xx loops, chains longer than 2 hops), and meta coverage (missing or duplicated title/description tags).

How often does the audit run?

Muginai runs a full site audit on every deploy and on a configurable schedule (daily or weekly). Critical issues — missing canonical tags, hreflang errors on high-traffic pages, or 5xx errors — trigger immediate Telegram alerts rather than waiting for the next scheduled report.

How are audit issues prioritized?

Each issue gets a priority score from 1 (cosmetic) to 5 (critical). The score factors in page traffic (from GSC if connected), issue type severity, and how many pages are affected. The audit panel always shows the highest-impact issues first.

🔬

Automated Site Audit

A 40-rule on-page audit engine that runs continuously — no Screaming Frog sessions, no manual crawl scheduling, no spreadsheet exports. Issues are prioritized, tracked, and delivered as Telegram alerts when critical.

40+ audit rules

P1–P5 priority scoring

Live Telegram alerts

What gets checked

🏷️

Structured data (schema.org)

JSON-LD presence and valid @type
Article, FAQPage, HowTo, LocalBusiness, Product schema validation
Required fields missing or malformed
Conflicting schema on the same page

🌍

hreflang & internationalization

hreflang tags present on multilingual pages
Return links verified (x→y requires y→x)
x-default set and pointing to correct URL
Locale mismatches between canonical and hreflang

⚡

Core Web Vitals

LCP > 2.5s flagged as P3 issue
CLS > 0.1 flagged on key landing pages
FID / INP monitoring via CrUX API (when configured)
Mobile vs desktop score comparison

🔗

Internal linking

Orphan pages — indexed but no inbound internal links
Anchor text diversity scoring
Deep pages (4+ click depth) surfaced for silo restructuring
Internal link count distribution

↩️

Redirect chains & errors

3xx redirect loops detected
Chains longer than 2 hops flagged for cleanup
404 pages linked from internal pages
5xx errors on crawled URLs

📋

Meta & on-page coverage

Missing or empty title / meta description
Duplicate title tags across pages
Duplicate meta descriptions
H1 missing or multiple H1 tags
Canonical tag missing, self-referencing, or cross-domain

Priority scoring — fix the right things first

P5 Critical 5xx errors on indexed pages · hreflang loop on homepage · missing canonical on key landing pages Telegram alert fires immediately

P4 High Schema validation failure · orphan pillar pages · redirect chain >2 hops on money pages Surfaces at top of audit panel

P3 Medium LCP >2.5s on landing pages · missing meta description · duplicate title tags Included in weekly digest

P2 Low Anchor text over-optimization · deep pages (3+ clicks) · internal link count outliers Logged for batch cleanup

P1 Info Trailing slashes · minor meta length deviation · subresource hints missing Available on demand

How the audit runs

Crawl trigger

Audit runs on every deploy and on schedule (daily or weekly). You can also trigger a manual audit from the panel or via Telegram /audit command.

Rule engine

40+ checks run against every crawled URL. Rules are weighted by page authority (GSC traffic data) and issue severity. New issues are diff'd against the previous run.

Priority scoring

Each issue is scored P1–P5 factoring in: issue type, page importance, and blast radius (how many pages are affected). Score informs sort order in the panel.

Alerts & reports

P4/P5 issues fire a Telegram alert immediately. All issues appear in the Audit panel grouped by category. Weekly digest includes net-new vs resolved issue counts.

Common questions

Do I need Screaming Frog or another crawler alongside Muginai? +

No. Muginai's built-in crawler handles URL discovery, fetch, and rule execution. For sites with 100,000+ pages, DataForSEO's site audit API can be connected as an alternative crawler — but for most projects, the built-in engine is sufficient.

How does hreflang validation work? +

Muginai crawls all pages in the sitemap, extracts hreflang annotations, and builds a bidirectional graph. It verifies that every x→y hreflang link has a corresponding y→x return link, that x-default is set, and that locale values are valid BCP 47 codes.

Can I mark issues as intentionally ignored? +

Yes. Any issue can be snoozed for 30/60/90 days or marked as permanently acknowledged. Snoozed issues don't re-appear in reports or alerts until the snooze period expires or you manually wake them.

Does it check JavaScript-rendered content? +

The default crawler uses HTTP fetch (fast, no JS rendering). For SPAs or pages requiring JavaScript execution, you can configure the Playwright-based crawler mode on Pro and Agency plans. JS rendering is slower but catches issues invisible to a standard crawler.

No more manual crawl sessions

Join the waitlist and get audit results from day one.

Get early access → All features View pricing