Documentation
How RankoraXPress works, how scores are calculated, and how to use the platform.
Scoring methodology
Every audited page contributes to seven site-wide scores: SEO, AEO, GEO, Technical Health, Content Quality, Schema Health, and AI Readiness. Each score is computed from named, weighted factors. For each factor, the score is the percentage of crawled pages that pass that check. The category score is the weighted average of its factors. The Search Visibility Score is a weighted composite (SEO 22%, AEO 16%, GEO 16%, Technical 16%, Content 10%, Schema 10%, AI 10%).
Click any score in the analyzer to expand the exact factors and weights that produced it. Scores are never inflated — they are earned through actual implementation.
Crawler architecture
When you enter a website, RankoraXPress validates the URL, fetches robots.txt, discovers the sitemap, parses sitemap URLs, then crawls eligible internal pages while respecting crawl depth, delay, and robots.txt disallow rules. Each page is normalized into a standard record (title, headings, canonical, links, images, schema, word count, response time) consumed by the SEO, AEO, GEO, technical, content, schema, and AI engines.
The analyzer offers three audit modes behind a single provider abstraction, so the seven analysis engines always consume identical normalized page data regardless of source:
- Quick Audit — in-browser crawl via CORS proxies. Always available; subject to browser cross-origin limits.
- Full Site Audit — server-side crawl job against a backend crawl-service. Crawls hundreds of pages without browser CORS limits. When no backend is configured it honestly reports “unavailable” and falls back — it never fabricates a server crawl.
- Demo Audit — synthetic sample data, always labeled “demo”, so the full UI is exercisable without a live crawl.
Because the Quick mode runs in the browser, page fetches go through a CORS proxy. If the proxy is blocked, the analyzer honestly reports a partial or demo result — it never claims to have crawled pages it could not fetch. A Full Site Audit against the backend removes this limitation and enables scheduled monitoring at scale.
Security: the crawler blocks private IP ranges, localhost, link-local, and cloud metadata endpoints to prevent SSRF — including every redirect destination. All user-provided URLs are validated and sanitized. The server provider applies the same SSRF guard at the DNS-resolution layer before connecting.
Backend crawl-service API contract
The Full Site Audit talks to a backend crawl-service over a small, documented HTTP API. Point RankoraXPress at your backend by setting localStorage["rankoraxpress.apiBase"] to the API origin (for example https://api.rankoraxpress.com). When unset or unreachable, Full Site Audit is honestly shown as unavailable.
| Method & path | Purpose | Response |
|---|---|---|
| POST /v1/crawl | Submit a crawl job. Body: { url, options }. Returns jobId and initial status. | job id + status |
| GET /v1/crawl/:id | Poll job status and progress until terminal (completed, partially_completed, cancelled, failed). | status + result |
| DELETE /v1/crawl/:id | Cancel a running job: stop scheduling new requests, abort active ones, preserve completed pages. | ack |
The terminal result must return the same normalized page-data model the frontend engines consume: URL, final URL, status, content type, title, meta description, H1/H2/H3, canonical, robots directives, word count, text content, internal/external links, images, alt attributes, structured data, Open Graph/Twitter metadata, response time, and response size. The backend must enforce SSRF at DNS resolution, per-request timeouts (15s), response-size caps (2.5MB), bounded concurrency (≤6), redirect-destination validation, authentication, authorization, job-ownership checks, and rate limiting.
The audit workflow
- Crawl — fetch the URL and parse the HTML.
- Understand — extract metadata, structure, schema, and content.
- Audit — run factor-level checks across six dimensions.
- Compare — benchmark against competitors and past audits.
- Prioritize — rank issues by severity and impact.
- Fix — apply the actionable recommendations.
- Validate — confirm the fix resolved the issue.
- Recrawl — re-audit to measure improvement.
- Monitor — schedule recrawls and get regression alerts.
Discovery files
- sitemap.xml — all indexable URLs.
- robots.txt — crawl directives for search and AI bots.
- llms.txt — machine-readable resource map for AI crawlers.
Feature docs
SEO Audit
How the SEO score is computed from on-page factors.
AEO Audit
Direct-answer patterns, FAQ, and answer engine extraction.
GEO Audit
Entity clarity and citation-worthiness scoring.
Technical SEO
Crawlability, indexability, and performance checks.
Schema Generator
Generate valid JSON-LD structured data.
SEO Monitoring
Scheduled recrawls and regression alerts.
Self-audit
RankoraXPress audits its own pages with the same engine it ships to customers. The platform's own scores target SEO 95+, AEO 90+, GEO 90+, Technical 95+, Content 90+, and AI Readiness 90+.
Audit your site in seconds
Run a free SEO, AEO, and GEO audit. See exactly what to fix and why.
No signup required · RankoraXPress