SCHEMA.md

Wiki Schema

Domain

AI in Education — research, products, policies, and pedagogical debates around the use of artificial intelligence (including LLMs, adaptive systems, tutoring agents, and analytics) in learning environments from K-12 to higher education and lifelong learning.

Conventions

• File names: lowercase, hyphens, no spaces (e.g., adaptive-learning-systems.md)
• Every wiki page starts with YAML frontmatter (see below)
• Use wikilinks to link between pages ([[page-slug]] or [[page-slug|display text]])
• Inline hyperlink rule (wiki-style, HARD GATE): hyperlink every concept mentioned by name in the BODY of a concept or article page to that concept's page. On an article page this inline link is the ONLY place the relationship is recorded (article pages have no Connected sections from 2026-09-27); concept, resource and FAQ pages keep their Connected Concepts/Articles lists as well. Use piped links when display text differs from the slug (e.g. [[cognitive-offloading|doing the cognitive work]]), the most specific slug matching the mention's meaning, and the dedicated umbrella page for generic terms (e.g. [[feedback]], not [[feedback-loop]], for plain "feedback"). This pass is a BLOCKING PREREQUISITE before npm run build / commit / push / deploy on every newly created or enriched page — a green build does NOT substitute for it. Load the wiki-inline-links skill and run the pass + verification (0 self-links, 0 heading links, balanced brackets, 0 broken links) on every new page first.
• When updating a page, always bump the updated date+time (see Frontmatter above)
• Every new page must satisfy the frontmatter contract of its collection (there is no page index to add it to since index.md was retired)
• Every action must be appended to log.md (local-only, gitignored like raw/ — not committed; persists on the local repo for scan-complete anchors)
• Three page types: content/en/articles/<slug>.md for individual papers, content/en/concepts/<slug>.md for broad topics that synthesize multiple papers, and content/en/faqs/<slug>.md for curated question-and-answer pages. An article belongs on a concept page's Connected Articles list; a concept page explains the concept itself, not any single paper; a FAQ answers a specific question and connects to concepts and articles via connected_faqs (below).
• Provenance markers: On pages that synthesize 3+ sources, append ^[raw/papers/source-file.md] at the end of paragraphs whose claims come from a specific source. This lets a reader trace each claim back without re-reading the whole raw file. Optional on single-source pages where the sources: frontmatter is enough.

Frontmatter

Only these fields are validated by the live schema (src/content.config.ts). Extra fields in frontmatter are ignored at build time, so keep to this list.

---
title: Page Title
created: "YYYY-MM-DDTHH:MM:SS±HH:MM"
updated: "YYYY-MM-DDTHH:MM:SS±HH:MM"
published: "YYYY-MM-DD"                # articles only, optional: the PAPER's publication date
type: article | concept | faq
confidence: high | medium | low        # how well-supported the claims are
sources: [raw/papers/source-name.md]   # articles only
connected_faqs: [faq-slug-1, faq-slug-2]  # concepts + articles only (optional)

# facet fields: concept slugs, each from its own registry section (a wrong-kind value fails the build)
foundations: [ai-literacy, cognitive-offloading]
pedagogy: [scaffolding, self-regulated-learning]
technology: [llm]
assessment: [formative-assessment]
methods: [quantitative-research]
institutions: [governance]
ethics: [trust-calibration]

# phrase fields: closed vocabularies, not concept slugs
research_method: [survey]
discipline: [cs education]
level: [higher ed]
audience: [instructors]
page_kind: [framework]
---

Concept references are TYPED (2026-09-17, replacing the old tags list). Every concept a page touches is named in one of the typed fields below, never in a free-floating list. Each facet field accepts only the concept slugs filed under its own section of concepts.registry.yaml, so a value of the wrong kind — a technology slug in pedagogy — is rejected by the build. If a topic isn't yet a concept, create the concept page first or fold it into the nearest existing concept; never invent a slug. Non-concept metadata (how the study was done, who it is for, which level, what kind of page) goes in the phrase-valued fields.

Structured metadata fields (optional, added 2026-08-29): research_method, discipline, audience, level, and page_kind hold non-concept metadata.

page_kind was called category until 2026-09-17 and held thirteen values, ten of which were topics: assessment, instructional design, engagement, equity, policy, curriculum design, ai foundations, collaborative learning, design thinking and well-being. Every one of those has a concept page, and the typed facet fields now carry them, so the topics were retired and the field was renamed to say what it actually holds: the kind of page this is - framework, synthesis or evaluation. A page that presents a framework is not a page about frameworks, and that distinction is the reason the field survives at all. These fields drive the search page filter facets and the Metadata table rendered at the foot of every page.

Typed facet fields (added 2026-09-17): foundations, pedagogy, technology, assessment, methods, institutions, ethics. A flat tag list mixed kinds together — a pedagogy, a technology, an outcome and an ethical theme in one list. Each facet field holds only concepts of one kind, so the search facets can ask "show me pedagogy studies" directly. Every facet mirrors one section of concepts.registry.yaml:

The UI label of every facet IS its registry section heading, read from the generated FACET_FIELDS in src/data/facetVocab.ts, so the sidebar, the search filters and the metadata table cannot call one section by three different names. The display order is generated too (FACET_DISPLAY_ORDER): page-level fields first, then the topics, with Foundations late, immediately before Page kind.

• foundations ← Foundations of AI in education
• pedagogy ← Learning and instruction (pedagogies, learning theories, learner processes, plus lifelong and workplace learning)
• technology ← Technologies and techniques (rendered Technologies and techniques; includes platforms, openness and technology adoption)
• assessment ← Assessment and measurement (assessment, feedback, measurement and validity)
• methods ← Research methods and evaluation (research methods, benchmarks, evaluation of AI systems)
• institutions ← Institutions and policy (policy, governance, regulation, change management, administrators)
• ethics ← Equity, ethics, and responsible use

Readership is not subject matter, and the labels say so. audience (rendered Intended audience) is who the page is written for - a role vocabulary of instructors, learners, researchers, curriculum designers, software developers - and holds no concept slugs. The subject-matter fields, institutions (Institutions and policy) among them, carry the concept slugs of their registry sections that the page analyses, such as AI governance and regulation.

The stakeholders field (People) was retired 2026-09-21. It restated audience on the pages that used both (174 pages read audience: [learners] beside stakeholders: [student-experience]), and its two most-used values were not people at all: student-experience is a learner-experience concept and teacher-ai-competency a capability, so the field had no definition that could be applied consistently. Readership is now the single question it answers, and audience answers it. The People section keeps its concept pages for browsing (like AI in the disciplines, it backs no field); teacher-ai-competency moved to Foundations of AI in education and student-experience to Learning and instruction, where they are usable as foundations and pedagogy values.

Method is two questions, not one, and only two. research_method (rendered Study design) is how the page's own study was carried out - survey, experiment, case study, instrument development, system development - and is a phrase field. methods (Research methods) is the set of method and evaluation CONCEPTS the page connects to: randomized controlled trials, design-based research, qualitative and quantitative research, meta-analyses, benchmarks, network analysis, latent profile analysis. The two are different granularities - this study versus the methods literature - which is why both exist and why their labels differ.

The phrase vocabulary used to duplicate the concept one, so ~300 pages stated their method twice in two vocabularies ("benchmark" in research_method AND the benchmark concept in methods). The duplicates were retired: benchmark, randomized controlled trial, meta-analysis, systematic review, design-based research, mixed methods, qualitative, quantitative, educational measurement, learning analytics and research methods are no longer study-design values, and the corresponding concept value is expected in its facet field instead. research_method is now what its label says: the design and the data collection.

Every metadata field label links to its umbrella concept page. Where a field's vocabulary has one page that reads as the umbrella for the whole field, the metadata table at the foot of a page links the field NAME there - Study design and Research methods both to the research methods page, Discipline to discipline-specific AI in education, Technologies and techniques to the AI technologies page. The map is generated (FIELD_UMBRELLAS in src/data/metadataLinks.ts) and the link is dropped if the page does not exist in the build, so a renamed or missing umbrella degrades to plain text rather than a dead link.

A value's wording is the taxonomy's wording (2026-09-20). educators folded into instructors and students into learners; faculty development became faculty developers, because the field names readers, not activities. librarians was added, arts education and music education cover the arts (the arts, design and media education concept is their page), environmental education covers sustainability, climate and green-skills work, and library and information science names the research field. level now runs as bands - preschool, primary education, middle school, secondary, k 12, higher ed, undergraduate, graduate, adult learning, special education, teacher education - with early childhood renamed to preschool, elementary to primary education and teacher training to teacher education. Values without a concept page yet are allowed in these fields; the concept pages follow the vocabulary, not the other way round.

The section AI in the disciplines (subject areas and research fields) is served by the hand-curated discipline field, and Levels and contexts by the level field, which also covers school subjects that have no concept page yet.

These fields are AUTHORED, then validated. Until 2026-09-17 they were projected from each page's tag list by derive-facets.py; with tags retired the fields are written directly and tooling/scripts/validate-facets.py enforces three things — every value is a concept filed under that field's section, no concept appears in two facet fields, and every page carries at least one typed value. It is a gate, so npm run verify fails on a violation. The allowed values are generated from the registry into src/data/facetVocab.ts, so adding a concept to a section extends the vocabulary on the next build.

tags was RETIRED on 2026-09-17. Every tag value had a typed home: 83% of the 10,968 tag entries were already concepts in the six facet sections, 15% were covered by discipline or level, and the remainder were migrated by hand. The field is gone from the schema, the page templates no longer render tag chips, and the JSON-LD keywords now come from the typed fields.

The Metadata table (2026-09-17) renders every typed field at the foot of each page, one row per field, with values hyperlinked to their concept page whenever one exists. Facet values are concept slugs and link directly; the phrase-valued fields resolve through src/data/metadataLinks.ts, generated from the registry so cs education reaches cs-education and systematic review reaches meta-analysis-systematic-review. The table states what a page IS. ## Connected Concepts remains the curated, relevance-ordered list of what a page relates to beyond that — the two are complementary, and a concept already named in the table does not need repeating in the list.

These are CLOSED vocabularies, not free-form text. The authoritative lists live in src/content.config.ts, and the Astro build rejects any page carrying a value outside them, so a typo fails the build rather than silently creating a one-page facet. The lists are reproduced below for convenience and checked against the schema by python3 tooling/scripts/audit-metadata.py --check-docs, which fails if this block drifts.

Each field is optional and each value must be chosen from its list. Omitting a field is the correct answer when nothing fits: a study with no disciplinary home should omit discipline rather than be forced into the nearest one, and an opinion piece may have no level. Use 1-3 values per field, most specific first.

research_method: action design research | bibliometric | case study | delphi | design and evaluation study | experiment | instrument development | interviews | latent profile analysis | literature review | longitudinal study | network analysis | policy analysis | position paper | process-outcome modeling | quasi-experiment | secondary analysis | structural equation modeling | survey | system development | thematic analysis | theoretical analysis | user study
discipline: arts education | biology education | business education | chemistry education | cs education | design education | engineering education | english education | environmental education | humanities education | information technology | language learning | legal education | learning sciences | library and information science | math education | medical education | music education | nursing education | physics education | science education | stem education | vocational education | writing education
audience: administrators | assessment designers | assessment professionals | curriculum designers | educational technology developers | faculty developers | institutions | instructional designers | instructors | learners | learning analytics designers | librarians | medical educators | parents and families | policymakers | researchers | software developers
level: adult learning | graduate | higher ed | k 12 | middle school | preschool | primary education | secondary | special education | teacher education | undergraduate
page_kind: framework | synthesis | evaluation
foundations: academic-integrity | agency | agentic-ai | ai-education | ai-literacy | cognitive-offloading | computational-thinking | critical-thinking | curriculum-design | design-thinking | educational-development | framing-ai-use-for-students | history-of-aied | human-ai-collaboration | interpreting-and-applying-aied-research | learner-identity | learning-design | limitations-in-aied-research | philosophy-of-ai-in-education | reducing-ai-misuse | samr-model | teacher-role | theories-and-frameworks | theory-development-aied | tpack
pedagogy: active-learning | activity-theory-aied | anxiety-and-stress | behaviorism | career-development-and-readiness | cognitive-psychology | collaborative-learning | community-of-inquiry | constructivist | creativity | critical-pedagogy | desirable-difficulties | distributed-cognition | embodied-learning | experiential-learning | game-based-learning | help-seeking | icap-framework | inquiry-based-learning | learning-by-teaching | learning-theories | lifelong-learning | mastery-learning | metacognition | misconceptions | motivation | online-teaching-and-learning | pedagogical-partnerships | pedagogy | prior-knowledge | problem-based-learning | problem-solving | productive-failure | professional-training | project-based-learning | refutation-text | retrieval-spacing-interleaving | scaffolding | self-determination-theory | self-directed-learning | self-efficacy | self-regulated-learning | situated-learning | social-emotional-learning | sociocultural-learning | socratic-method | storytelling-in-education | student-ai-interaction | student-engagement | transfer-of-learning | video-education | well-being
technology: adaptive-learning | affective-computing | affective-tutoring | ai-technologies | cognitive-diagnosis | conversational-ai | edtech-platform | educational-nlp | educational-robotics | generative-ai | human-in-the-loop-ai | intelligent-tutoring | knowledge-graph | knowledge-tracing | learning-analytics | llm | machine-learning | multimodal | open-source | pedagogical-agent | llm-training-and-fine-tuning | personalized-learning | prompt-engineering | rag | recommender-systems-and-learning-paths | reinforcement-learning | simulating-students | simulation | speech-and-voice-technologies | student-modeling | technology-acceptance-model | vibe-coding | virtual-and-augmented-reality | visualization
assessment: ai-detection | ai-feedback-quality | assessment | assessment-validity | authentic-assessment | automated-assessment | automated-essay-scoring | automated-question-generation | educational-measurement | eportfolio | evaluative-judgment | feedback | feedback-literacy | formative-assessment | group-work | item-response-theory | learning-gains | peer-assessment | psychometrically-aware-ai | remote-proctoring | self-report-measures | summative-assessment
methods: ai-ed-evaluation | benchmark | design-based-research | latent-profile-analysis | meta-analysis-systematic-review | mixed-methods-research | network-analysis | qualitative-research | quantitative-research | rct | research-methods-aied | usability-research
audience: administrators | assessment designers | assessment professionals | curriculum designers | educational technology developers | faculty developers | institutions | instructional designers | instructors | learners | learning analytics designers | librarians | medical educators | parents and families | policymakers | researchers | software developers
institutions: change-management | educational-policy-ai | governance | regulation
ethics: accessibility | ai-misuse-learning-harm | ai-sycophancy | ai-use-disclosure | assistive-technology | bias-mitigation | culturally-relevant-pedagogy | differential-effects-across-learner-groups | digital-divide | equity-in-ai-education | ethics | explainable-ai | global-south | guardrails | hallucination-risk | inclusive-learning | legal-issues-and-risks | multilingual-learning | neurodiversity | pedagogical-safety | privacy | sustainability | trust | trust-calibration | universal-design-for-learning

Check the current coverage and any violation without a full build:

python3 tooling/scripts/audit-metadata.py             # coverage per collection + violations
python3 tooling/scripts/audit-metadata.py --strict     # non-zero exit on a violation
python3 tooling/scripts/audit-metadata.py --missing research_method

research_method, audience and level are expected on article pages unless the article genuinely gives no basis; discipline and category are expected only where they clearly apply. Adding a new value is a deliberate act: add it to src/content.config.ts first, then update the block above, and prefer extending the list over inventing a near-duplicate of an existing value.

connected_faqs (concepts and articles only, optional) lists FAQ slugs the page should link to in a Connected FAQs section at the bottom of the page. The section renders only when at least one listed FAQ exists. FAQs are wiki-linked from concept/article narratives and from other FAQs like any other page (inline [[wikilink]]), and a FAQ's own narrative can link to concepts, articles, and other FAQs.

created / updated MUST store full quoted date+time timestamps (e.g. "2026-08-16T20:47:13-04:00"), never bare dates. Reasons:

• The right sidebar ("Recently Added Articles" / "Recently Updated Concepts") and RSS sort by these fields via string comparison — date-only values tie within a day and fall back to alphabetical order. Full timestamps give correct reverse-chronological ordering.
• YAML parses an unquoted ISO timestamp into a JS Date in UTC, shifting an Eastern-evening value to the next calendar day. Always quote the value so the schema preserves the original string. tooling/scripts/check-frontmatter-dates.py (gate 11) enforces this: it fails on any date field that YAML has auto-typed, and on created/updated that are not full quoted timestamps with an offset. The damage it guards against is a vault editor (Obsidian and similar) re-serialising a frontmatter block and dropping the quotes — harmless-looking, but it moves the page's calendar day for the journal, sidebar and RSS.
• created should be the wiki ingestion date (with time), NOT the paper's publication date — Recent Articles and the journal sort by it. The paper's own date goes in published.
• published (articles only, optional) records when the paper was published, so a page can be dated honestly: an older paper ingested today shows published 2025-11-04 · added 2026-09-20 in the page header, and the JSON-LD ScholarlyArticle carries the paper's real date instead of the ingestion year. Precision follows the source: YYYY-MM-DD when the PDF or arXiv line gives a day, YYYY-MM for a month-only journal issue, YYYY when only the year is known — never invent a day. It is a plain string (bare dates are fine and preferred); any other shape fails the build. It is NOT used for sorting, so leave it off a page rather than guessing.
• When you make a significant body edit to a page (not just frontmatter or Connected Articles/Concepts lists), bump updated to the current date+time and rebuild so the sidebar refreshes.

Pages display date-only everywhere (article/concept page headers, sidebar) via .split('T')[0]; the time is stored internally for sorting only.

confidence is optional (defaults to medium) but recommended for opinion-heavy or fast-moving topics. A confidence: low page signals weak or partial support so weak claims don't silently harden into accepted wiki fact. There is no contested or contradictions field — the schema does not validate them, so do not add them to frontmatter. Record unresolved disagreements in the page body instead (see Update Policy below).

Article page body structure

> **Synthesis:** 150-220 words: what the paper does, how, and what it finds

## Key Findings
- 5-7 contiguous bullets, one finding each (no blank lines between them)

## <2-4 body sections>
Named for this paper's content, not a template

## What this means for practice
- Lead with **Instructors.** Add a labeled bullet for designers, administrators or researchers
  only when the paper supports a genuinely distinct implication for them.

## Limitations            (optional: evidence-bound limits only)
- 2-6 bullets, each tied to a concrete fact about this study

## Connected FAQs
- [[faq-slug]]      (optional; appended by the page template, not authored)

## Citation
Author, A. (2026). [Title](https://doi.org/...). Journal, Vol(Issue), pages.

Article pages carry NO ## Connected Concepts / ## Connected Articles section (rule change 2026-09-27): cross-link related concepts and articles as inline [[wikilinks]] in the body prose. The curated Connected lists live on concept, resource and FAQ pages. ## Citation is always the LAST section on the page. Exactly one per article, as a single APA-style line: the paper's title is the only hyperlinked text, first 6 authors + ", et al." for longer lists. The body carries no separate PDF/DOI link. Article body budget: 750-1,750 words, read from wiki.config.yaml (article.min_words/max_words) (frontmatter to ## Citation), and ## Limitations is written only when the study supplies real limits — a boilerplate "small sample, single institution" line is a defect, not a section.

Concept page body structure

> **Synthesis:** definition + overview of the concept

## Questions to Consider      (REQUIRED on every concept page)
- Pre-reading questions that make the reader reason, not recall

## Introduction

## <body sections with [[wikilinks]]>

## Connected Concepts
- [[concept-slug]]

## Connected Articles
- [[article-slug]]

## Connected FAQs             (optional)

(No citation section — concepts synthesize multiple sources.)

FAQ page body structure (content/en/faqs/<slug>.md)

---
title: "Question being answered?"
created: "YYYY-MM-DDTHH:MM:SS±HH:MM"
updated: "YYYY-MM-DDTHH:MM:SS±HH:MM"
type: faq
foundations: [ai-literacy]             # facet fields work on FAQ pages too
audience: [instructors]
---

# Question being answered?

Narrative answer with [[wikilinks]] to concepts, articles, and other FAQs
(mirror the aggressive inline-link convention of articles and concepts).

FAQ pages have no Connected Concepts/Connected Articles/Citation sections and no sources field (they are curated answers, not paper summaries). They are indexed in llms.txt/llms-full.txt, appear on the journal page (❓ badge), and are listed in concept/article "Connected FAQs" sections via connected_faqs. Numbered lists inside a FAQ must be written as ONE contiguous block (no blank lines between items) so they don't render as repeated 1. (see the list-formatting HARD GATE).

Resource page structure (content/en/resources/<slug>.md)

---
title: "Pressing Prompts"
created: "YYYY-MM-DDTHH:MM:SS±HH:MM"
updated: "YYYY-MM-DDTHH:MM:SS±HH:MM"
type: resource
summary: "One sentence for the /resources listing and the exports."   # REQUIRED
url: https://pressingprompts.org/                                     # REQUIRED, http(s)
source_code: https://github.com/owner/repo                            # optional
author: "Hannah Rozear, Remi Kalir and Aria Chernik"                  # REQUIRED, a person or a project
author_url: https://pressingprompts.org/about                         # optional
resource_type: [collection of activities]                             # REQUIRED (one or more)
access: [free]                                                        # free | free with account | freemium
license: "CC BY-NC-SA 4.0"                                            # optional, only when the site states one
last_verified: "YYYY-MM-DD"                                            # REQUIRED, date the link was confirmed
connected_resources: [pause-ai-use-self-examination]                  # existing resource slugs only
foundations: [ai-literacy]                                            # facet/level/audience/discipline as on any page
audience: [instructors]
level: [higher ed]
---

> One-paragraph lede: what it is and why it matters.

## What it is
## What you can do with it            (subsections are free-form; no fixed set)
## Who it is for
## Notes and caveats                    (optional: ownership, data handling, limits the project states)
## Connected Concepts

[[critical-pedagogy]], [[inquiry-based-learning]], [[ai-literacy]]

Resource pages catalog free tools, collections, instruments and open formats a reader can go and use — not papers. They therefore have no sources, no research_method, no page_kind, no Citation section and no "What this means for practice"/"Limitations" pair: there is no study to summarize and no claim to ground. They do keep the typed facets, level, audience and discipline, so a resource joins the same concept graph and the same search filters as everything else.

Requirements and conventions:

• url is the point of the page. Never leave it empty, and never point it at a page that was not actually opened — link rot is the failure mode here, which is why last_verified is mandatory.
• author is a real person or the project itself. When a project publishes no author, name the project (or the GitHub account) and say so in the body; never invent a person.
• resource_type values: software, ai tutor, agent skill, prompt or gem library, collection of tools, collection of activities, assessment instrument, open format or specification, ebook or guide, case study collection, dataset or benchmark.
• access describes what a reader has to give up to use it; license only records a license the site itself states.
• connected_resources is frontmatter (validated against content/en/resources/), so the section can never point at a page that does not exist. The same field is available on articles, concepts and FAQs.
• Encyclopedia-style body, roughly 200–350 words across the sections above (a platform with several capability areas sits at the top of that range), US English, [[wikilinks]] to concepts in the body and a ## Connected Concepts list naming the concepts the resource serves.
• Rendered at /aied/resources/<slug>/; listed on the /aied/resources/ index (grouped by first resource_type), as the closing "Free Tools and Resources" chapter of the EPUB/PDF exports.

raw/ Frontmatter

Raw sources ALSO get a small frontmatter block so re-ingests can detect drift:

---
source_url: https://example.com/article   # original URL, if applicable
ingested: YYYY-MM-DD                       # or ingested_date
sha256: <hex digest of the raw content below the frontmatter>
---

raw/ is gitignored and never committed. Raw files may also carry an authors: list (used as the source of truth when a page's citation authors need verification). The sha256: lets a future re-ingest of the same URL skip processing when content is unchanged, and flag drift when it has changed. Compute over the body only (everything after the closing ---), not the frontmatter itself.

Concept taxonomy

The taxonomy lives in concepts.registry.yaml, the single source of truth: every concept is filed under one of seven sections and their groups, and that filing decides which facet field accepts it. The old tag taxonomy that used to be reproduced here was retired with the tags field — see tooling/concept-index.md for the generated human-readable manifest.

Page Thresholds

• Create a page when an entity/concept appears in 2+ sources OR is central to one source
• Add to existing page when a source mentions something already covered
• DON'T create a page for passing mentions, minor details, or things outside the domain
• Split a page when it exceeds ~200 lines — break into sub-topics with cross-links
• Archive a page when its content is fully superseded — remove it from the live site

Concept Pages

One page per concept or topic. Include:

• Definition / explanation
• Current state of knowledge
• Open questions or debates
• Related concepts (wikilinks)

Ingestion enrichment rule: when a new article makes a significant contribution to a connected concept (a novel framing, distinctive finding, or a dimension the concept page lacks), integrate that insight into the concept page's body narrative — a research bullet, subsection, or synthesis paragraph — not merely add it to the Connected Articles list. Evaluate each article's substantive contribution and incorporate it where the concept is missing that dimension. When such a significant edit is made, bump the concept's updated timestamp and rebuild so the right sidebar reflects it.

Update Policy

When new information conflicts with existing content:

1. Check the dates — newer sources generally supersede older ones
2. If genuinely contradictory, note both positions with dates and sources in the page body
3. If the contradiction is significant, lower the page's confidence to medium or low
4. Flag for user review in the lint report

Journal

• The dated journal is a site page (/journal), built from the content collections grouped by each page's created date, newest first.
• There is no journal.md: it was retired on 2026-09-27 rather than kept in sync with the page that readers actually see.
• Shows: type badge (📄 article / 🏷️ concept / ❓ FAQ), wikilink to page, source reference, full title, and tags.
• Excludes low-confidence stubs with no sources.
• Regenerate after every ingestion batch to keep current.
• FAQs appear with a ❓ badge; concepts (that aren't low-confidence stubs) appear with a 🏷️ badge.