BE DISCOVERABLE BY AI
robots.txt, sitemaps, canonicals, indexability and AI crawler access form the discovery foundation.
Six-tier evidence hierarchy.
We never conflate an official standard, a vendor guideline, an evolving proposal, and an internal heuristic.
OFFICIAL_STANDARD
OFFICIAL_STANDARD
RFCs (e.g. RFC 9309 robots.txt), W3C standards, Schema.org official vocabularies.
OFFICIAL_VENDOR
OFFICIAL_VENDOR
Published vendor crawler documentation from OpenAI, Anthropic, Google, and Perplexity.
PROPOSAL
PROPOSAL
Emerging specifications such as llms.txt v2. Evaluated with limited weight and marked as evolving.
MEASURED
MEASURED
Directly observed HTTP responses, status codes, headers, and SSL parameters.
INTERNAL_HEURISTIC
INTERNAL_HEURISTIC
Internal algorithmic patterns detecting dead-end navigation or anti-patterns.
EXPERIMENTAL
EXPERIMENTAL
Early-stage agent discovery signals (e.g. MCP endpoints), labeled separately without score inflation.
Unmeasured signals: The NOT_MEASURED contract.
Core Web Vitals cannot be fabricated from raw HTML fetches. Without real user CrUX metrics, performance hygiene strictly reports NOT_MEASURED.
SSRF GUARD
Private IPv4/IPv6 ranges, loopback addresses, and cloud metadata endpoints fail closed before any TCP handshake to prevent SSRF vulnerabilities.
AI SEO · CHATGPT VISIBILITY · ENGINE V3
See which technical blocker limits discoverability before you spend time fixing the wrong thing.
HTML&HTML audits public website surfaces, shows evidence for each finding, and prioritizes a fix path your technical team can execute.
Diagnosis is free.Actionable fixes, acceptance checks and rollback guidance unlock with the one-time $99 pack.
Decision area
FREE AI SEO AUDIT
$99 IMPLEMENTATION PACK
Deterministic audit
18 Engine V3 modules · 105 controls · no random scoring
Finding → root cause → fix → test → rollback
Crawl and security
Up to 50 public HTML pages · 30 live link probes · fail-closed SSRF
Evidence-bound implementation plan
AI search disciplines
SEO · GEO · AEO · LLMO · AAO · RAG · E-E-A-T
Evidence-bound code and configuration
Decision intelligence
13 non-scoring analyses · 7 readiness lenses · NOT_MEASURED / REQUIRES_CONTEXT
P0–P3 priority, dependencies, acceptance and regression
AI visibility observation
Bot access, source readiness, entity/schema and answer extractability
When provider keys are configured, up to 3 neutral queries on API/search-grounded surfaces
Release safety
Evidence and confidence classes remain visible
G0–G9 · acceptance · regression · rollback
Evidence boundary: External-model recommendation, ranking, citation, traffic or revenue is not guaranteed. Unavailable evidence stays NOT_MEASURED; code context stays REQUIRES_CONTEXT.
Paid delivery boundary: up to 30 evidence-bound page-level Markdown machine surfaces and a versioned ZIP delivery package.