Chaos
Machine Trust Intelligence

From digital chaos
to machine intelligence.

SiteNexis transforms fragmented web signals into structured machine understanding. 12 intelligence scores. 16 autonomous agents. Real-time retrieval simulation. See exactly where AI systems trust your content — and where they don't.

Free  ·  No account required  ·  Results in ~60s

12 intelligence scores6-stage retrieval simulationInvestigation-grade diagnostics
Scroll to discover

The Gap

AI systems don't read pages.
They retrieve, rank, and cite chunks.

Your content is stripped of navigation, split into 300–600 token semantic units, embedded, ranked against competing chunks, compressed into a generated answer, and filtered for citation eligibility. That is six failure points before your name appears in an AI response.

Traditional tools audit what humans see. SiteNexis audits what machines extract — running 16 autonomous agents across four dependency layers to model the complete AI retrieval and trust formation pipeline.

Screaming Frog
Ahrefs
SEMrush
PageSpeed
Schema.org
Surfer SEO

SiteNexis

Intelligence Modules

Four intelligence layers.
All in one scan.

Layer 1

SEO Intelligence

Title, meta, canonicals, robots, sitemaps, broken links, crawlability signals.

Layer 2

AI Readability

Chunk quality, entity clarity, conversational readiness, summarisability.

Layer 2

Schema Engine

Schema detection, field validation, auto-generated snippets, structured trust.

Layer 1

Link Graph

Internal link topology, PageRank distribution, orphaned pages, anchor quality.

Layer 2

Content Intelligence

Topical depth, entity coverage, citation readiness, semantic trust signals.

Layer 1

Performance

Core Web Vitals, LCP, CLS, INP, TTFB, mobile performance benchmarks.

Layer 4 — Machine TrustPro+

Retrieval Simulation · Machine Trust · Temporal Authority

Models how AI retrieval systems form, maintain, and decay trust in your content — from chunk extraction to recommendation surface presence.

Unlock Layer 4

Surface Coverage

Four AI surfaces.
One coverage map.

Retrieval and recommendation are not the same thing. A page can score well on AI Visibility and still be absent from AI Overviews, chat responses, voice answers, and agent discovery. SiteNexis maps your inclusion probability across all four surfaces — and diagnoses exactly what is blocking each one.

Google AI Overviews82
ChatGPT (browsing)67
Perplexity74
Claude (web search)91
Gemini58

All surface scores are estimated from measurable content signals.

Intelligence Dashboard

One scan. Twelve scores.
Complete explainability.

SiteNexis
Complete16 agents · 142 pages
74
AI Visibility+12
61
Entity Trust+8
88
Retrieval Ready+3
52
Machine Trust+21
79
Citation Score+6
Critical Issues3 critical
criticalPrimary entity definition missing across 47 pagesAI Visibility −18pts
criticalNo sameAs validation — external trust signals absentMachine Trust −15pts
warning12 pages lack structured data — retrieval riskRetrieval Score −8pts

Live AI Audit Observatory

Watch real websites
earn trust, live.

These are not mockups. Every score below comes from a real crawl, every issue from live evidence, every recommendation is deterministic. Expand a card to see the evidence, replay how it was produced, or open the full report.

Every audit above is real — run by SiteNexis against a live domain. Scores, issues, and recommendations are generated from that crawl, not a template. Run your own audit above to see the same depth of evidence.

Intelligence Scoring

Twelve scores.
Every one explainable.

Every deduction maps to a named, actionable issue. No composite score without a sub-score breakdown. No black boxes.

1
Crawl & Structure
2
Semantic Intelligence
3
AI Visibility
4
Machine Trust
View full score breakdown

example.com

Last audited 2 min ago

74
AI Visibility
+1274
Entity Trust
+861
Retrieval Ready
+388
Machine Trust
+2152
Citation Score
+679
12 issues detected3 critical

Methodology

How SiteNexis works.
Every step, every agent, every score.

01

Full-Site Crawl

The Crawl Agent renders every page with Puppeteer, strips navigation and boilerplate, extracts semantic chunks of 300–600 tokens, and pre-identifies entities. Up to 500 pages per audit. Respects robots.txt.

02

Parallel Layer 1–2 Analysis

Six agents run simultaneously: SEO (title, meta, canonicals, sitemaps), Schema (structured data validation), AI Readability (chunk quality, extractability), Entity Intelligence (consistency, disambiguation), Citation Probability (7-factor weighted score), and Semantic Trust (contradiction detection via Claude API on top 20 pages).

03

Layer 3 AI Visibility Scoring

AI Visibility Score = Machine Readability × 15% + Entity Confidence × 20% + Retrieval Readiness × 20% + Citation Probability × 20% + Semantic Trust × 15% + Schema Completeness × 10%. The AI Perception Graph is built from entity relationships and typed connections. Every deduction maps to a named, actionable Issue.

04

Layer 4 Machine Trust (Pro+)

Five Layer 4 agents run in parallel after Layers 1–3 complete. Retrieval Simulation models 6 stages on top 30 pages by PageRank. Machine Trust scores entity credibility, schema alignment, external validation, contradiction absence, and decay signals. Temporal Authority tracks velocity and drift across audits. Recommendation Mapping scores inclusion probability across four AI surfaces. Synthetic Entity Detection identifies manufactured authority patterns.

05

Information Gain + Scout

The Information Gain Engine fetches the SERP cohort via Serper API and measures what your content adds that the top results do not already cover — entity gaps, question gaps, evidence uniqueness. Scout classifies every page by query intent (informational, commercial, research, voice, agent) and scores structural alignment between intent and page signals.

06

Global Fix Plan + Reports

The Fix Plan aggregates every issue from all 12 modules into a single P0/P1/P2 queue with cross-module dependency chains, effort estimation, and three-dimension impact scoring (SEO / AI Visibility / Trust). The Reporting Agent generates a full PDF with all scores, issues, and recommendations. Every API route returns a GTL state envelope: complete, partial, or empty.

Pricing

Start free.
Scale with intelligence.

Free

$0

One audit per month.

  • 1 audit / month
  • All Layer 1–2 modules
  • PDF export
  • No credit card
Start Free
Starter

$29/mo

For independent developers.

  • 50 audits / month
  • All Layer 1–3 modules
  • PDF reports
  • Email support
Get Started
ProMost Popular

$79/mo

Full intelligence stack.

  • Unlimited audits
  • All 4 intelligence layers
  • Machine Trust (Layer 4)
  • Competitive analysis
  • Priority support
Start Trial
Agency

$249/mo

For managed client portfolios.

  • Unlimited audits
  • Bulk domain scanning
  • White-label reports
  • API access
  • Dedicated manager
Contact Sales

Product Suite

Two platforms.
One intelligence suite.

SiteNexis models how AI systems retrieve and trust your content. AdNexis models what creative drives conversion when they do. Together, they cover the full signal-to-conversion pipeline.

SiteNexis

AI Retrieval & Machine Trust Intelligence

You are here

Models how AI systems retrieve, interpret, trust, and recommend your website — across all four intelligence layers from crawl structure to machine trust formation.

  • Machine Trust Score — 5-dimension trust model
  • Entity Intelligence — consistency + disambiguation
  • Retrieval Simulation — 6-stage pipeline modeling
  • Recommendation Surfaces — AI Overviews, chat, voice, agents

AdNexis

AI Creative Intelligence

New

Deconstructs top-performing ads with AI to extract what makes them convert — then generates platform-specific creative variations in seconds.

  • Hook Intelligence — classify, score, and rank ad hooks
  • Full Ad Analysis — funnel stage, CTA, emotion, audience
  • Swipe Vault — your AI-analysed ad library
  • Creative Generation — variations for Meta, TikTok, YouTube

Machine Resource Studio

Turn crawl data into clear decisions.

Machine Resource Studio explains what the crawl found, why the finding matters to Google and AI systems, and which fix should come first. Each score shows its formula, evidence count, confidence, and limits.

Explore the studio

Evidence first

Recommendations link to crawl pages or issue records.

Explainable scores

Every score card shows the formula and confidence.

Google and AI views

Read technical impact and machine interpretation impact separately.

Honest limits

Missing data stays unavailable. It does not become zero.

Get Started Free

See what AI systems
actually extract from your site.

16 agents. 12 intelligence scores. A Global Fix Plan with P0/P1/P2 priorities. One audit — free, no account required.

Layer 4 Machine Trust analysis requires Pro plan or above.