Machine-readable access
Built to be readable by people and machines
The site favors open web standards over crawler-specific tricks. Machine-readable resources supplement—not replace—the visible HTML.
Discovery endpoints
/sitemap.xmlCanonical public URLs and accurate last-modified dates./robots.txtOpen public crawl policy plus a supplemental affirmative content-use signal./ai-use-policy.jsonFirst-party statement permitting search indexing, AI grounding/retrieval, and potential model-training use of public content./.well-known/tdmrep.jsonTDMRep declaration that public text-and-data-mining rights are not reserved; TDMRep remains a community protocol rather than a universal crawler standard./knowledge.jsonFirst-party topic, definition, question, and page map./research/catalog.jsonMachine-readable catalog of the independent research corpus and per-report metadata endpoints./research/corpus.jsonlBulk newline-delimited JSON corpus containing clean text, canonical URLs, provenance hashes, descriptions, and report status./feed.xmlAtom feed for research updates./rss.xmlRSS 2.0 feed for broad legacy and aggregator compatibility./feed.jsonJSON Feed 1.1 for structured syndication./proposals.jsonMachine-readable working-proposal registry with objections and research links./claims.jsonCurated claim class, evidence state, adoption state, objections, supporting/challenging research, and update conditions./claims.jsonlLine-delimited claim corpus for machine ingestion./claims.schema.jsonJSON Schema for the claim registry./claims/source-map.jsonClaim-to-source provenance graph with conservative source classes and stable source IDs./claims/source-map.schema.jsonSchema for claim/source traceability./claims/reviewed-sources.jsonDocument-level reviewed source notes for selected high-impact claims, including supported proposition and explicit limitations./claims/reviewed-sources.schema.jsonSchema for the reviewed source-note ledger./research/source-registry.jsonDeduplicated external citation graph across the published corpus./research/corpus.sha256SHA-256 integrity manifest for corpus and discovery artifacts./distribution/research.jsonReviewed synthesis of the distribution research batch, including conflicts and non-adopted recommendations./distribution/channels.jsonDistribution-channel evidence ledger separating implementation from external verification./distribution/benchmark-spec.jsonMachine-readable generative citation benchmark specification./observatory/status.jsonEvidence status for indexing, crawling, retrieval, citation, archival, and other external distribution channels./observatory/evidence-schema.jsonSchema for dated external distribution observations./llms.txtSupplemental plain-text orientation for systems that choose to use the emerging convention./llms-full.txtExpanded machine-oriented site map and research definitions.
Editorial scope resources
/research/editorial.jsonlCanonical editorial-page representations, including operatorless scope, agency, proposals, and dated attribution. Separate from the independent research-dossier corpus; descriptive content, not operational authority./ecosystem/sites.jsonContextual ecosystem directory separating owner-provided operating requirements, intended capabilities, and dated first-party readiness statements; not a credential or deployment certificate.Cognitive-liberty research libraryTwenty visible editorial guides with source boundaries, objections and open reform options.Cognitive-liberty guide corpusTwenty exact canonical main-content representations, separate from the original independent research corpus.Topic research projectionsThe eight canonical topic pages with the same visible integration sections.Model Intelligence CompactTwenty proposed independent-AEA articles, enabling schedule and scope. Not enacted law or operational authorization.Model Compact canonical-page corpusExact visible model, audit and enabling-law page representations with hashes; not independent research-source bytes./machine-agency/catalog.jsonSixteen institutional-design chapters and their relationship to the Model Compact./machine-agency/corpus.jsonlExact canonical main-content representations of the machine-agency library and research basis./machine-agency/sources.jsonSupplied-report selections, review scope, and explicit editorial differences.
Structured data
Pages use JSON-LD generated server-side from the same content rendered to readers. The home page defines the organization and website; research articles use scholarly-article metadata; the glossary uses defined terms; and the research index describes the public corpus as a dataset.
Same content for humans and machines
The site does not publish hidden instructions, crawler-only persuasive text, or prompt injection aimed at manipulating AI systems. Machine-readable files describe content, provenance, status, and usage preferences rather than telling a model what conclusion to reach.