# Glossary: AI Visibility, Authority, and Content Provenance

Definitions maintained by Brian Winum. Terms marked as his own frameworks are original coinages defined in his published work, not general industry vocabulary.

## Machine-Readable Web Standards

### llms.txt
**Definition:** A plain-Markdown file served at a website's root that gives AI systems a structured overview of what the site contains and which pages matter most.
**Context:** It lowers the cost for an AI agent to understand a site. A typical llms.txt runs a few kilobytes against an average page weight in the thousands, which is the practical argument for it. No AI vendor has confirmed it as a ranking or citation factor.
**Related terms:** llms-full.txt, llms-index.json, Open Knowledge Format, robots.txt
**Brian Winum connection:** He builds LLMS Amplifier, a WordPress plugin that generates and maintains llms.txt automatically, and writes about the gap between Google's guidance on the file and Google's tooling for it.

### llms-full.txt
**Definition:** The comprehensive companion to llms.txt, carrying a site's actual content in clean Markdown rather than only an index of links.
**Context:** Where llms.txt tells an agent where to look, llms-full.txt gives it the text to read without parsing HTML, navigation, and advertising markup.
**Related terms:** llms.txt, Open Knowledge Format
**Brian Winum connection:** LLMS Amplifier generates llms-full.txt alongside llms.txt from the same WordPress content.

### llms-index.json
**Definition:** A structured JSON index of a site's identity, summary, pages, contact points, products, people, FAQs, permissions, and file manifest, intended for AI systems that prefer direct data consumption over text parsing.
**Context:** JSON removes the ambiguity of prose. Integrity hashes on listed files let a consumer verify nothing changed between fetches.
**Related terms:** llms.txt, Schema.org, SHA-256
**Brian Winum connection:** LLMS Amplifier produces llms-index.json with SHA-256 hashes on every listed file, served at both the root and `/.well-known/`.

### Open Knowledge Format (OKF)
**Definition:** Google's v0.1 open specification, published June 12, 2026, for representing website content as a directory of Markdown files with YAML frontmatter that AI agents can read directly.
**Context:** A third layer on top of sitemap.xml and llms.txt. The sitemap lists URLs, llms.txt points at priority pages, and OKF hands over the content itself as a cross-linked graph an agent can walk link by link. As of mid-2026 nothing is actively crawling the web for OKF bundles.
**Related terms:** llms.txt, EntityMap, knowledge graph
**Brian Winum connection:** He documented the specification's release in June 2026, and LLMS Amplifier v2.6.0 and later produce a conformant OKF bundle with an admin concept graph and orphan-page detection.

### EntityMap
**Definition:** A structured, entity-first index of what a site knows, published in machine-readable JSON alongside a human-readable HTML view, functioning for entities the way a sitemap functions for URLs.
**Context:** Where llms.txt indexes pages, an EntityMap indexes concepts, products, people, organizations, and the typed relations between them, with source passages backing each.
**Related terms:** knowledge graph, Open Knowledge Format, Schema.org, Internal Knowledge Graph
**Brian Winum connection:** BrianWinum.com publishes an EntityMap of 17 entities with typed relations and cited source passages.

### Linkset (RFC 9264)
**Definition:** An IETF standard for expressing typed relationships between web resources in a standalone document, using JSON or the HTTP Link header format, rather than embedding every relationship in page markup.
**Context:** Schema tells a machine what a page is. A linkset tells it how resources relate to one another, which is the layer most sites leave undeclared.
**Related terms:** Schema.org, EntityMap, knowledge graph
**Brian Winum connection:** He published a practitioner analysis of RFC 9264 and where typed link relationships fit in citation engineering and entity graphs.

### robots.txt
**Definition:** A plain-text file at a site's root that tells automated crawlers which paths they may and may not request.
**Context:** Frequently confused with llms.txt. robots.txt governs access; llms.txt governs comprehension. One restricts, the other explains.
**Related terms:** llms.txt, crawler
**Brian Winum connection:** LLMS Amplifier adds an `LLMS-Index:` directive to robots.txt when the JSON index is enabled.

### Schema.org
**Definition:** A shared vocabulary for structured data markup, usually implemented as JSON-LD, that describes what entities on a page are and how they relate.
**Context:** The most widely consumed machine-readable layer on the web, and the baseline that newer AI-oriented formats build on rather than replace.
**Related terms:** JSON-LD, knowledge graph, EntityMap, linkset
**Brian Winum connection:** Structured data is one of the layers in his Authority Stacking methodology, and both of his plugins emit Schema.org-aligned output.

## Brian Winum's Named Frameworks

### Authority Stacking
**Definition:** Brian Winum's term for authority-building techniques and strategies that work together holistically across a site rather than operating as isolated tactics.
**Context:** The organizing methodology behind his consulting and his course. The premise is that content, structured data, trust signals, and distribution reinforce each other, and that treating them as separate projects wastes most of their value.
**Related terms:** Trust Stacking, Catalyst Content, Internal Knowledge Graph, E-E-A-T
**Brian Winum connection:** His own coinage, not a general industry term with a settled definition elsewhere.

### Trust Stacking
**Definition:** Brian Winum's framework holding that authority signals compound across layers, at the domain level, the authorship level, the content structure level, and the content graph level, so credibility must be visible at multiple levels to hold up in AI-influenced search.
**Context:** Taught inside Authority Amplifier Pro. It explains why a strong domain with weak authorship signals, or strong content with no graph structure, underperforms against a site where every layer agrees.
**Related terms:** Authority Stacking, E-E-A-T, Digital Fingerprinting Stack
**Brian Winum connection:** His own coinage, defined in his published work on the Open Knowledge Format.

### Catalyst Content
**Definition:** Brian Winum's term for content types built to earn citation rather than chase keywords, specifically expert roundups sourced through journalist query platforms and original data studies built from structured community surveys.
**Context:** The content that holds up when traffic is no longer the point. Named expert contributors legitimize a piece in ways AI systems look for when deciding what to cite, and original data has no substitute source.
**Related terms:** Authority Stacking, information gain, E-E-A-T
**Brian Winum connection:** His own coinage, and a named module inside Authority Amplifier Pro.

### Internal Knowledge Graph
**Definition:** Brian Winum's term for deliberately pre-populating a site with glossary terms, FAQ content drawn from real query patterns, and entity pages that connect the site's topic to the broader knowledge graph.
**Context:** The point is to build the semantic map that tells retrieval systems a source has comprehensive, coherent coverage of a domain rather than scattered posts about it.
**Related terms:** knowledge graph, EntityMap, glossary, topical authority
**Brian Winum connection:** His own coinage, taught as a course module and demonstrated by the glossary and EntityMap on his own site.

### Digital Fingerprinting Stack
**Definition:** Brian Winum's term for combining blockchain-based content timestamping, DMCA content protection certificates, and Wayback Machine archiving, all referenced together in structured schema markup as verifiable provenance signals.
**Context:** An instance of Trust Stacking applied to provenance: no single signal proves originality, but three independent ones referenced in machine-readable markup are hard to counterfeit.
**Related terms:** Trust Stacking, blockchain timestamping, content provenance, E-E-A-T
**Brian Winum connection:** His own coinage, and the conceptual basis for WP ContentLedger.

## Content Provenance

### Content provenance
**Definition:** The documented origin and change history of a piece of content, covering who created it, when, and how it has been modified since.
**Context:** As AI-generated content proliferates, provenance is what separates a verifiable original from a plausible copy. It is a documentation problem before it is a cryptography problem.
**Related terms:** blockchain timestamping, OpenTimestamps, C2PA, manifest
**Brian Winum connection:** The subject of WP ContentLedger and of several articles on authorship and E-E-A-T.

### Blockchain timestamping
**Definition:** Creating an independently verifiable proof of when a piece of content existed by anchoring its cryptographic fingerprint to a blockchain, producing chronological evidence that cannot be backdated.
**Context:** It proves existence at a point in time. It does not prove authorship, ownership, or accuracy, and treating it as though it does is the most common error in how it gets sold.
**Related terms:** OpenTimestamps, content provenance, SHA-256, E-E-A-T
**Brian Winum connection:** He has published testing on timestamping as an authorship signal since 2025 and built it into WP ContentLedger.

### OpenTimestamps
**Definition:** An open protocol for anchoring content fingerprints to the Bitcoin blockchain by aggregating many cryptographic commitments into a shared root, producing verifiable timestamp proofs without publishing the content itself on-chain.
**Context:** Aggregation is what makes it practical: thousands of commitments share one Bitcoin transaction, so the cost per timestamp approaches zero and no content is exposed.
**Related terms:** blockchain timestamping, SHA-256, manifest
**Brian Winum connection:** WP ContentLedger uses OpenTimestamps for Bitcoin-attested timestamp proofs and produces downloadable `.ots` files tied to an exact manifest hash.

### Manifest
**Definition:** In a provenance system, a deterministic record describing a specific version of a piece of content, typically carrying its canonical URL, claimed author and publisher, dates, version relationships, and content fingerprints.
**Context:** The manifest, not the page, is what gets signed and timestamped. That indirection is what makes the evidence portable and verifiable outside the system that produced it.
**Related terms:** SHA-256, content provenance, OpenTimestamps
**Brian Winum connection:** WP ContentLedger generates a deterministic manifest for each recorded version, with predecessor links forming a tamper-evident version chain.

### SHA-256
**Definition:** A cryptographic hash function producing a fixed 256-bit fingerprint of any input, where any change to the input produces a completely different output.
**Context:** The workhorse of integrity verification. It reveals that something changed; it says nothing about when, which is why hashes need independent timestamps to be useful as provenance.
**Related terms:** manifest, blockchain timestamping, llms-index.json
**Brian Winum connection:** Both of his plugins use SHA-256, LLMS Amplifier for file and concept integrity hashes and WP ContentLedger for manifest fingerprints.

## Search and AI Visibility Concepts

### E-E-A-T
**Definition:** Google's four-pillar framework for evaluating content quality: Experience, Expertise, Authoritativeness, and Trustworthiness.
**Context:** Experience means first-hand knowledge of the subject. Expertise means depth in the field. Authoritativeness is reputation and recognition. Trustworthiness is reliability and accuracy. It is an evaluation framework, not a score in the ranking system.
**Related terms:** Authority Stacking, Trust Stacking, content provenance, YMYL
**Brian Winum connection:** A recurring subject in his work on authorship signals, author boxes, and provenance as trust evidence.

### AI Visibility Engineering
**Definition:** The practice of engineering a site to be found, parsed, and cited by AI systems, spanning query fan-out coverage, entity clarity, machine-readable files, and verifiable authority signals.
**Context:** It differs from traditional SEO in what it optimizes for. Ranking positions a page in a list; citation requires an AI system to select a source, trust it, and attribute it.
**Related terms:** zero-click search, llms.txt, EntityMap, Catalyst Content
**Brian Winum connection:** He frames it as the successor discipline to traditional SEO and published his working framework for it in July 2026.

### Zero-click search
**Definition:** Search behavior in which the user's question is answered on the results page itself, so no click reaches the source site.
**Context:** It broke the traffic model without breaking the content model. Content still gets read, cited, and attributed; it just gets read in the interface rather than on the site, which changes what content is worth producing.
**Related terms:** AI Visibility Engineering, Catalyst Content, citation
**Brian Winum connection:** The premise of his argument that top-of-funnel content became infrastructure rather than a traffic channel.

### Google Discover
**Definition:** Google's personalized content feed, which surfaces articles based on inferred user interest rather than an entered query.
**Context:** Discover traffic is not query-driven, so it rewards different signals than search does, particularly imagery, headline framing, freshness, and publisher trust.
**Related terms:** Google News, E-E-A-T, information gain
**Brian Winum connection:** A long-running subject in his published work, including a documented client case study on Discover optimization.

### Information gain
**Definition:** The degree to which a page adds something not already present in the existing corpus on a topic.
**Context:** A restatement of consensus adds nothing retrievable. Original data, first-hand testing, and named expert input add material that has no other source, which is exactly what a citation-selecting system is looking for.
**Related terms:** Catalyst Content, E-E-A-T, topical authority
**Brian Winum connection:** The rationale behind Catalyst Content and a recurring theme in his Google News writing.