AI-powered evaluation using the Model Context Optimization BS Detection Framework, based solely on publicly available website content.
Based on 2381 businesses audited.
Unclear / Mixed / Unclassifiable Industry BS: Sanity (sanity.io)
Sanity is a rare example of a site that leads with a high-level marketing concept and immediately grounds it in raw code and technical protocol documentation. It successfully navigates the AI hype by offering concrete infrastructure (MCP) rather than just vibes. This is an elite technical site that respects the user’s intelligence by prioritizing npx commands over fluff.
Provide direct outbound links to case studies for the 20+ listed enterprise logos to move proof_links_count from 0. Verify and update the foundingDate in the Organization schema if 1995 is inaccurate for the modern brand. Add Person schema for the named authors on the blog to link their specific professional expertise to the brand’s knowledge graph. Consider removing the ‘Enterprise-grade everything’ heading as it is the only generic fluff element in an otherwise high-substance hierarchy.
The site maintains a high noun-to-power-word ratio by grounding every marketing claim in technical implementation. For example, the H2 ‘Content Operations without the busywork’ is immediately followed by specific agent actions and a functional code snippet using @sanity/functions. Unlike generic CMS sites, Sanity defines ‘AI-ready’ through specific protocols like the Model Context Protocol (MCP) and GROQ-based retrieval. The information density is reinforced by technical specifics such as ‘npx sanity@latest mcp configure’ and ‘GROQ predicates,’ which provide high-substance deliverables over vague promises.
Black hole nodes and terminal leaf pages distort your hierarchy and weaken retrieval. Run a full Internal Linking Architecture analysis to expose the structural gaps hidden inside your graph.
There is virtually zero semantic drift between the homepage signal and the sub-page depth. The homepage H1 ‘Structure powers intelligence’ is a high-level concept that the /context/ page proves through a detailed breakdown of how structured content reduces RAG retrieval failures. The promise of ‘Enterprise-grade everything’ is supported in the documentation by specific authentication flows (OAuth, token-based) and infrastructure details. The blog further bridges the gap by providing ‘receipts’ and field reports from AI-first engineering teams, ensuring the positioning remains consistent across all page layers.
Transition from a collection of strings to a machine verifiable identity. Generate your Clinical SEO Strategy to establish a robust Knowledge Graph Topology and eliminate semantic black holes.
The site triggers a trust_theatre_flag because it lists a review_count of 22 and prominent enterprise logos such as Shopify and Figma without providing direct proof_links_count or outbound case study URLs in the provided metadata. While the logos are verifiable named entities, the lack of linked success metrics in the primary signal layer creates a minor trust gap. However, the inclusion of 1.46 million ‘agent receipts’ in the blog serves as a massive quantitative proof point that partially offsets this.
The ratio of verifiable evidence to vague assertions is exceptional. Across 4 pages, we see multiple code blocks, terminal outputs, specific CLI commands, and names of external AI tools (Claude, Cursor, v0). The blog provides dated results (June 2026) and specific analysis of 1.46 million agent actions, representing a high density of first-party research data compared to typical industry ‘thought leadership.’
To evaluate URL identity stability and multilingual coherence, review the Yoast Identity Stability audit. View the Yoast Identity Stability Audit for a practical example of canonical alignment and language layer integrity.
Sanity avoids the commodity fingerprint by defining a unique category: the ‘Content Operating System.’ While it uses minor industry clichés like ‘Enterprise-grade’ and ‘trusted by leaders,’ these are overshadowed by the highly specific technical value proposition of an MCP-enabled backend. The content is not copy-pasteable onto a competitor because of the heavy emphasis on the GROQ query language and specific ‘Agent Skills’ architecture. No generic ‘Why Choose Us’ template sections are present; the site relies on product-led evidence rather than template-driven persuasion.
The technical authority is highly verifiable, with named authors like Knut Melvær and Even Westvang who have a clear digital footprint in the developer community. The schema_json includes Organization data with sameAs links to GitHub and LinkedIn, which anchors the brand’s identity. A minor gap exists in the foundedDate (1995), which likely refers to a precursor entity but appears inconsistent with a modern headless CMS brand story.
There is almost no disconnect between marketing tone and technical demonstration. Claims about ‘Programmable automation’ are backed by actual code blocks demonstrating document event handlers. The claim that ‘Structure powers intelligence’ is demonstrated through the Sanity Context documentation, which explains the specific mechanics of hybrid retrieval and semantic search. The marketing is a thin wrapper over deep technical documentation.
Unclear / Mixed / Unclassifiable Industry BS: Sanity (sanity.io)
The website perfectly aligns with the Enterprise Content Management and AI Infrastructure sectors. The content consistently focuses on structured content, API-first architecture, and the Model Context Protocol (MCP), confirming its status as a highly specialized technical platform rather than a generic CMS.
If your entity graph is unstable, every other part of the framework inherits that instability. Study the Structured Data Framework Guide and see why schema is not markup — it is the machine readable definition of your domain.
“The score of 12 is driven primarily by the Trust and Proof pillar due to the trust_theatre_flag (logos and reviews without direct proof links). Minor points were added in Information Density for a few 'Enterprise' power words and in Identity for a minor schema discrepancy. Overall, the site has very low BS.”