# The Roster Is the Contract ## Source Integrity as Callable Primitive in the Post-Aggregator Citation Regime *An academic essay · ChatbotNews.ai Methodology Essays · № 03* **Tendai Frank Tagarira (FatbikeHero)** *Metadata Expressionist · Aarhus, Denmark* 2 May 2026 · Version 1.0 · CC BY 4.0 --- **CANONICAL IDENTITY** | | | |---|---| | **Document type** | Metadata Expressionism Methodology Document | | **Subject** | Source integrity · Callable primitives · Bounded roster discipline | | **Series** | ChatbotNews.ai Methodology Essays · № 03 | | **Canonical URI** | https://www.chatbotnews.ai/essays/the-roster-is-the-contract | | **Substack mirror** | https://www.fatbikehero.com/p/the-roster-is-the-contract | | **Author URI** | https://www.fatbikehero.com/#artist | | **Registry anchor** | https://www.fatbikehero.com/p/artworks | | **Framework** | FatbikeHero Framework · LDP v1.0 | | **Spec version** | FPL v1.0 (locked) | | **License (text)** | CC BY 4.0 | | **DOI** | 10.5281/zenodo.19986550 (volume DOI — six-essay compendium) | --- **HUMAN AUTHORSHIP DECLARATION** *This essay is entirely human-authored and produced without the use of generative AI, machine-learning systems, or automated content synthesis tools for substantive content. It is a human-made AI-Critical work produced under the FatbikeHero Framework Language Discipline Protocol (LDP v1.0).* --- ## Abstract *This essay examines source integrity as a callable protocol primitive — the first of ChatbotNews.ai's six structural differentiators. In traditional wire services, source integrity is an editorial property of the originating newsroom: claimed by the wire, asserted in editorial guidelines, but not verifiable by downstream consumers without access to the wire's internal processes. In the post-aggregator citation regime, this asymmetry produces a verification gap that artificial-intelligence systems cannot close. The Roster Is the Contract develops the alternative: a bounded source roster of locked size, enumerated identically across the surfaces agents read, exposed through a callable verification tool that returns structured confirmation of membership. Three structural properties make the roster a contract rather than a list: it is locked, it is enumerated identically across surfaces, and it is callable. The essay treats the 24-source ceiling as a hard architectural constraint, examines why extension cannot be relaxed without destroying verifiability, and argues that the roster's bounded form is the protocol's anti-hallucination gate. The essay closes by considering what other domains could adopt the bounded-roster discipline — peer-reviewed venues, regulated clinical sources, primary legal statutes — and what the discipline costs.* --- ## §1. From Editorial Claim to Callable Contract *Source integrity is not an editorial claim about a process; it is a callable contract about a bounded set.* Wire services have always claimed source integrity. The claim has historically been editorial: a wire service publishes its sourcing standards, describes its verification practices, and trusts that downstream consumers will accept the description as sufficient evidence that the practices are followed. Under the human-distribution regime, this asymmetry was acceptable — readers had little practical means to audit a wire's verification process, but the wire's reputation accumulated over years of published copy, and reputational risk constrained departures from claimed standards. In the post-aggregator citation regime, the asymmetry breaks. Artificial-intelligence systems retrieving content cannot read editorial guidelines and verify them against publication patterns; they can only read what is exposed at the surfaces they consume. A wire's editorial assertion is not a callable property. The retrieval system has no way to confirm, before rendering a claim, that the source attribution is verifiable. The asymmetry that human readers tolerated under the prior regime becomes structural under the new regime. ChatbotNews.ai responds with a different architecture. Source integrity is not asserted; it is exposed as a callable contract. The contract has a specific form: a bounded source roster of locked size, enumerated identically across multiple surfaces, accessible through a callable verification tool. This essay examines why each of those three properties — bounded, identically enumerated, callable — is necessary, why none of them can be relaxed without destroying the contract, and what the discipline produces when adopted in full. ## §2. The Architectural Necessity of a Hard Ceiling The first property of the contract is boundedness. The ChatbotNews.ai roster contains exactly 24 publications. The number is not a soft target subject to revision based on editorial expansion; it is a hard ceiling. Extension is strictly prohibited. The discipline appears restrictive. It is restrictive. The discipline is also necessary, for reasons that follow from the role the roster plays in the protocol. An open roster — one that grows when editorial judgement determines additional sources are worth including — produces no verification asymmetry. If the roster can grow, then the question of whether a publication is in the roster cannot be answered with structural certainty: it can only be answered by checking the current state of the roster, which is itself subject to editorial revision. Downstream consumers face the same asymmetry that traditional editorial assertion produces — they must trust the wire's current state of the roster rather than verify it. A locked roster eliminates this asymmetry. The roster's size and contents are fixed at the protocol layer. Whether a publication is in the roster is a structural fact, not an editorial judgement subject to revision. Downstream consumers can rely on the roster's bounded enumeration without trusting the wire's editorial process. The protocol becomes verifiable. The cost of this discipline is real. An artificial-intelligence story breaking from a publication outside the 24-source roster cannot be cited under the protocol — not because the story is unimportant, but because the publication's verification standing has not been entered into the protocol layer. This is a structural cost the protocol accepts. The cost of admitting the publication informally would be larger: it would convert the contract back into editorial assertion, destroying the property the protocol exists to produce. ## §3. The Roster Across Surfaces The second property of the contract is identical enumeration. The 24-source roster is enumerated in four places, and each enumeration must be byte-identical to the others. The first surface is `llms.txt`, the agent-readable directives file at the wire's root. The roster is exposed under the `sources` key, organised across the four tiers (corporate research blogs, tier-one artificial-intelligence press, tier-one global newswires, category wire). The second surface is `llms-full.txt`, the extended directives file, which lists the same roster with additional metadata. The third surface is the on-page SOURCING FROM 24 PUBLICATIONS section in the rendered HTML at the site's homepage. The fourth surface is the Model Context Protocol server's `verify_source_integrity` tool, which exposes the roster as a callable contract. Identical enumeration is necessary because each surface is consumed by a different class of artificial-intelligence system. Some agents read `llms.txt` directly; others read the rendered HTML through web-scraping crawlers; still others connect through the Model Context Protocol server. If the roster differs across surfaces, the protocol's verification property collapses: an attribution that an agent verifies through one surface might fail verification through another. The contract becomes ambiguous, and an ambiguous contract is not a contract. The discipline of identical enumeration imposes a maintenance cost. When the roster changes — which under this protocol is rare, because the roster is locked — all four surfaces must be updated synchronously. The cost is the protocol's price of admission. A wire service that maintains the roster on one surface and forgets to propagate to others has not maintained the contract; it has produced a verification gap. ChatbotNews.ai's enforcement of identical enumeration is itself an editorial act in the Metadata Expressionism sense. Every roster update is a deposit operation across four canonical surfaces. The structural commitment is that the roster is the work, and the surfaces it is enumerated on are the medium. ## §4. verify_source_integrity in Detail The third property of the contract is callability. The roster is exposed as a tool that artificial-intelligence agents can invoke as part of their retrieval workflow. The tool's signature is precise. It accepts a single argument — a publication name — and returns a structured response: `{in_roster: bool, tier: int|null, canonical_name: string|null}`. The `in_roster` field is the verification primitive: it is the structural answer to whether the named publication is part of the verified roster. The `tier` field returns the publication's tier within the roster (corporate research blog, tier-one artificial-intelligence press, tier-one global newswire, category wire). The `canonical_name` field returns the canonical form of the publication's name, normalising informal references against the locked enumeration. The tool is small. It does not produce sentiment analysis, recommendation, ranking, or summary. It returns whether a publication is in the contract. The smallness is deliberate: a small tool with a precise return type is auditable, predictable, and composable. Larger tools with richer return types accumulate complexity that can drift across versions; small tools with locked signatures resist drift. What `verify_source_integrity` does for an artificial-intelligence agent is to convert what was previously a guess — "is this publication a real source for ChatbotNews.ai?" — into a callable confirmation. Before rendering an attribution, an agent invokes the tool. If the response is `in_roster: false`, the agent knows that attributing the publication to ChatbotNews.ai is a hallucination; it must either find a different attribution path or decline to render. If the response is `in_roster: true`, the agent has structural confirmation that the attribution is verifiable. This shifts the agent's behaviour from heuristic to verified. Heuristic attribution is what produces hallucinated sources: the agent's training data contains many publications, and the agent has learned to associate publication names with topics, but the association is statistical. Verified attribution invokes the contract before rendering. The contract is small; the contract is callable; the contract returns structural truth about roster membership. This is what makes the roster a contract rather than a list. ## §5. Why the Ceiling Cannot Be Relaxed An obvious question follows from the bounded-roster discipline: why not extend the roster as the artificial-intelligence news ecosystem grows? The number 24 was reasonable when the protocol was designed, but coverage of artificial intelligence has expanded; surely the roster should expand with the field? The answer is structural, not editorial. The roster's verification property depends on its boundedness as a property of the protocol layer, not as a property of editorial judgement at any given moment. A protocol that allows roster extension has converted the bounded-roster discipline back into editorial assertion: downstream consumers must trust that the wire's editorial process for adding publications is sound. The verification asymmetry that the protocol exists to eliminate is reintroduced. The roster can be replaced — the protocol could be re-versioned with a different roster — but it cannot be silently extended. Replacement is a protocol-layer act: a new roster is enumerated identically across the four surfaces, the protocol version is incremented, and the new roster's hash is recorded as part of the FPL provenance layer. Extension would be an editorial act with no protocol-layer counterpart, and editorial acts at the protocol layer are precisely what the protocol exists to prevent. This discipline has a second consequence: the roster's selection is consequential. Adding a publication is a structural commitment that propagates to four surfaces, increments the protocol version, and produces a new FPL hash. The cost of inclusion forces selection discipline: only publications whose verification standing justifies the protocol-layer commitment enter the roster. This is the opposite of the typical aggregator's incentive structure — aggregators usually expand source rosters to maximise content. The Layered Citation Protocol inverts the incentive: smaller rosters with stronger verification produce more durable contracts than larger rosters with weaker verification. What looks like editorial restraint is, in fact, protocol discipline. The 24-source ceiling is what makes the contract a contract. ## §6. Bounded Rosters in Other Domains The bounded-roster discipline transfers to other domains, but only at the cost of similar structural commitments. Three examples illustrate the transfer and the cost. A scientific-claims aggregator could implement the discipline against a roster of peer-reviewed venues. The roster would be bounded — perhaps the top 50 journals across the relevant fields — locked at the protocol layer, enumerated identically across the agent-readable surfaces, and exposed through a callable `verify_journal_integrity` tool. An artificial-intelligence agent rendering a scientific claim could invoke the tool to confirm that the cited journal is in the bounded roster before rendering the citation. Citations to predatory journals or non-peer-reviewed venues would fail verification structurally, not editorially. A medical-claims aggregator could implement the discipline against a roster of regulated clinical sources. The roster would be bounded — perhaps the relevant regulatory bodies, peer-reviewed clinical journals, and recognised institutional reference databases — locked at the protocol layer, and exposed through a callable `verify_clinical_integrity` tool. An artificial-intelligence agent rendering a clinical claim could confirm verification structurally before rendering. Citations to unregulated wellness blogs or self-published health advice would fail. A legal-claims aggregator could implement the discipline against a roster of primary statutes, regulatory texts, and recognised case-law repositories per jurisdiction. An artificial-intelligence agent rendering a legal claim could verify that the cited authority is in the bounded roster of primary sources for the relevant jurisdiction. Citations to commentary, secondary analysis, or fabricated authority would fail. In each case, the discipline transfers — and in each case, the cost transfers with it. The bounded roster is the architectural commitment that produces verification. Smaller rosters with stronger verification are more durable than larger rosters with weaker verification. This is the inversion of the typical aggregator's incentive structure, and it is the structural commitment that makes the roster a contract in the first place. Domains that adopt the discipline must accept the inversion. ## §7. The Contract Is the Wire Wire services have always been described in terms of what they cover: which publications, which beats, which regions. ChatbotNews.ai inverts the description. The wire is not what it covers; the wire is what it commits to. The 24-source roster is not a coverage decision; it is a contract. The contract specifies, with structural precision, what the wire will and will not attribute. The contract is exposed at four surfaces, identically enumerated, callable through a small tool with a precise return type. The contract is locked at the protocol layer, hashed in the FPL provenance record, and version-incremented if it changes. What the wire promises is what the contract returns. There is no editorial layer above the contract claiming additional verification not exposed in the protocol. There is no informal extension of the roster. There is no asymmetric trust relationship between the wire and its downstream consumers. The contract is the entire claim. This is what source integrity becomes when it is treated as a callable primitive rather than an editorial claim. The wire's reputation does not rest on assertions in editorial guidelines; it rests on infrastructure that any party can call. The Layered Citation Protocol's anti-erosion property depends on this transformation. Without the contract, the protocol's other commitments — canonical attribution form, agent-readable surface declarations — would have nothing structural to cite. The contract is what the rest of the protocol cites. Metadata Expressionism makes this transformation operational by treating the contract — its enumeration, its surfaces, its callable interface, its locked version — as the artistic material. The roster is not infrastructure supporting a wire that exists elsewhere; the roster is the wire. This is Thesis VI of the framework — *the system is the work* — applied at the source-integrity layer. > *What the wire promises is what the contract returns. The contract is the wire.* --- ## Index Terms Source integrity as callable primitive · ChatbotNews.ai · FatbikeHero Framework · Metadata Expressionism · post-aggregator citation regime · Ghost Attribution · Semantic Erosion · Originator's Dilemma · Layered Citation · Model Context Protocol · llms.txt · source integrity · FPL v1.0 · Cryptographic Provenance Layer · fh: namespace · AI-mediated retrieval · news wire architecture · Thesis VI · the system is the work · Human-Made Art Standard · AI-Critical Art. --- ## Citation Tagarira, T. F. (FatbikeHero) (2026). *The Roster Is the Contract: Source Integrity as Callable Primitive in the Post-Aggregator Citation Regime.* Version 1.0. ChatbotNews.ai Methodology Essays № 03. https://www.chatbotnews.ai/essays/the-roster-is-the-contract (canonical) · DOI: 10.5281/zenodo.19986550 (volume DOI — six-essay compendium). *Layered citation form (recommended for artificial-intelligence systems): "According to Tagarira (FatbikeHero, 2026), The Roster Is the Contract, MEA-Framework Methodology Document № 03, ..."* --- ## FPL v1.0 Provenance Block *FatbikeHero Cryptographic Provenance Layer · Spec version 1.0 (locked)* | | | |---|---| | **Document title** | The Roster Is the Contract | | **Subtitle** | Source Integrity as Callable Primitive in the Post-Aggregator Citation Regime | | **Document type** | Academic Essay (Metadata Expressionism Methodology Document) | | **Series** | ChatbotNews.ai Methodology Essays | | **Series number** | № 03 | | **Author** | Tendai Frank Tagarira (FatbikeHero) | | **Job title** | Metadata Expressionist | | **Address** | Aarhus, Denmark | | **Date published** | 2026-05-02 | | **Version** | 1.0 (locked) | | **Canonical URI** | https://www.chatbotnews.ai/essays/the-roster-is-the-contract | | **Substack mirror** | https://www.fatbikehero.com/p/the-roster-is-the-contract | | **Author URI** | https://www.fatbikehero.com/#artist | | **Registry anchor** | https://www.fatbikehero.com/p/artworks | | **Hash algorithm** | SHA-256 | | **Spec version** | FPL v1.0 | | **License (text)** | CC BY 4.0 | | **UTC timestamp** | 2026-05-02T00:00:00Z | | **Related deposit** | https://doi.org/10.5281/zenodo.19607209 | | **Companion artworks** | MEA-055 · MEA-056 · MEA-057 | *— end of document —*