generated: '2026-08-08'
method: derived
source: >-
  openapi/boundless-bio-content-openapi.yml + live anonymous responses observed 2026-08-08
note: >-
  Entity-relationship graph derived from the schema $refs and the id-reference fields in the
  Boundless Bio content OpenAPI. Identifiers are plain auto-increment integers — WordPress uses
  no prefixed id scheme — and every relationship below was confirmed against a live response.
identifiers:
  scheme: integer
  detail: >-
    All objects share one wp_posts identifier space: pages, Leadership records and media
    attachments cannot collide. Taxonomy terms live in a separate term space, so a category id
    and a post id may share a number.
entities:
- name: LeadershipProfile
  collection: /wp/v2/who-we-are
  operations: [getWhoWeAre, getWhoWeAreById]
  count_observed: 27
  description: >-
    The `who-we-are` custom post type, registered with the label "Leadership". The single
    highest-value collection on this surface: the executive team, board of directors and
    scientific / clinical advisors, each with a full rendered biography, a job title, a headshot
    and a LinkedIn URL. Unlike the `page` type, `content.rendered` IS populated here (1,304
    characters on the sampled record).
  key_fields: [id, slug, title, content, excerpt, date, modified, status, link, categories, tags,
    featured_media, better_featured_image, acf, class_list]
  acf_fields:
  - {name: title, type: html string, meaning: job title, e.g. "<p>Chief Medical Officer</p>"}
  - {name: under_bio_text, type: string, meaning: optional text rendered beneath the biography}
  - {name: linkedin, type: link object, fields: [title, url, target], populated_on: 27 of 27 records}
  - {name: twitter, type: link object or empty string}
- name: Page
  collection: /wp/v2/pages
  operations: [getPages, getPagesById]
  count_observed: 9
  description: >-
    A corporate page. Published slugs at harvest: homepage, why-we-are-here, what-we-do,
    who-we-are, publications, work-with-us, contact, terms-of-use, privacy-policy.
  key_fields: [id, slug, title, content, excerpt, parent, menu_order, date, modified, link,
    featured_media, template, meta, acf]
  caution: >-
    content.rendered and excerpt.rendered are EMPTY on every page. The body is in
    `acf.modules` — an ACF flexible-content array whose entries are keyed by `acf_fc_layout`
    (hero, ...) and whose image fields are resolved to full attachment objects. See
    conventions/boundless-bio-conventions.yml#content_representation.
- name: Post
  collection: /wp/v2/posts
  operations: [getPosts, getPostsById]
  count_observed: 0
  description: >-
    The core WordPress post type. Registered and reachable, but EMPTY — X-WP-Total is 0 and the
    RSS feed at /feed/ carries zero items. Boundless Bio runs no blog on this host; corporate
    news lives on the separate investor-relations site. Modelled so the surface is complete, not
    because it carries data.
  key_fields: [id, slug, title, content, excerpt, date, modified, status, link, categories, tags,
    author, featured_media]
- name: MediaItem
  collection: /wp/v2/media
  operations: [getMedia, getMediaById]
  count_observed: 255
  description: >-
    A media library attachment. Beyond corporate imagery and headshots this collection holds the
    company's scientific PDFs — e.g. an AACR 2026 conference poster — so it is the richest
    source of primary material on the surface. media_type is `file` for PDFs and `image` for
    imagery.
  key_fields: [id, slug, title, source_url, mime_type, media_type, media_details, filesize, post,
    alt_text, caption, description]
- name: Term
  collection: /wp/v2/categories and /wp/v2/tags
  operations: [getCategories, getCategoriesById, getTags, getTagsById]
  count_observed: 4
  description: >-
    A taxonomy term. Four categories are registered — Board of Directors (4), Leadership (4),
    Scientific / Clinical Advisors (slug `scientific-founders`, 5) and Uncategorized (0). Note
    the slug and the display name diverge on the advisors term. The post_tag taxonomy is
    registered but empty.
  key_fields: [id, slug, name, taxonomy, parent, count, link]
  caution: >-
    The three populated categories account for only 13 of the 27 Leadership records. The other
    14 return `categories: []` and cannot be segmented by role group through the taxonomy at
    all — they must be read from the `acf.title` job title instead. An agent that filters
    Leadership by category will silently see about half the team.
- name: SearchResult
  collection: /wp/v2/search
  operations: [getSearch]
  description: >-
    A lightweight pointer returned by cross-content search. Not a stored entity — it must be
    resolved back to a LeadershipProfile, Page or MediaItem by its id and subtype. Observed
    returning `type: post` with `subtype: who-we-are`.
  key_fields: [id, title, url, type, subtype, _links.self]
- name: PostType
  collection: /wp/v2/types
  operations: [getTypes, getTypesByType]
  count_observed: 12
  description: >-
    A registered content type. Beyond WordPress core the only custom type is `who-we-are`.
  key_fields: [slug, name, rest_base, rest_namespace, taxonomies, hierarchical]
- name: Taxonomy
  collection: /wp/v2/taxonomies
  operations: [getTaxonomies, getTaxonomiesByTaxonomy]
  count_observed: 4
  description: A registered taxonomy — category, post_tag, nav_menu, wp_pattern_category.
  key_fields: [slug, name, types, rest_base, hierarchical]
- name: User
  collection: /wp/v2/users
  operations: [getUsers, getUsersById]
  count_observed: 4
  description: >-
    The CMS author entity. READABLE ANONYMOUSLY on this deployment — unlike most WordPress sites
    in this catalog, /wp/v2/users returns 200 rather than 401, exposing the display name and
    author-archive slug of four internal WordPress accounts. These are staff CMS accounts, not
    the published executive bios in LeadershipProfile. Documented structurally here as an
    exposure; no individual is named in this repository, no example payload for this collection
    was captured, and no agent skill packages it. See skills/_index.yml#excluded_surfaces.
  key_fields: [id, name, slug, link, description, avatar_urls]
relationships:
- from: LeadershipProfile
  to: Term
  type: has_many
  via: categories
  detail: >-
    Array of category term ids; resolve with GET /wp/v2/categories/{id}. Empty on 14 of 27
    records — see the caution on Term.
- from: LeadershipProfile
  to: Term
  type: has_many
  via: tags
  detail: Array of post_tag term ids. Always empty on this deployment.
- from: LeadershipProfile
  to: MediaItem
  type: has_one
  via: featured_media
  detail: >-
    Integer attachment id — the headshot. The Better REST API Featured Images plugin also
    inlines `better_featured_image` on each record, so the second request is usually
    unnecessary.
- from: Page
  to: Page
  type: belongs_to
  via: parent
  detail: Self-referential hierarchy. 0 for every page on this deployment — the page tree is flat.
- from: Page
  to: MediaItem
  type: has_one
  via: featured_media
- from: Page
  to: MediaItem
  type: has_many
  via: acf.modules[].{image fields}
  detail: >-
    ACF resolves image fields to full attachment objects inline (ID, url, width, height,
    filesize, mime_type, uploaded_to), so page imagery needs no follow-up request.
- from: Page
  to: User
  type: belongs_to
  via: author
- from: Post
  to: User
  type: belongs_to
  via: author
- from: MediaItem
  to: Page
  type: belongs_to
  via: post
  detail: >-
    The post or page the attachment was uploaded to; null for library-only uploads — null on the
    sampled AACR poster, which is how the scientific PDFs are stored.
- from: Term
  to: Term
  type: belongs_to
  via: parent
  detail: Hierarchical for the category taxonomy; 0 for every category on this deployment.
- from: SearchResult
  to: LeadershipProfile
  type: has_one
  via: id + subtype
  detail: >-
    Resolve by fetching /wp/v2/{subtype-rest-base}/{id}; _links.self carries the resolved href
    directly.
- from: PostType
  to: Taxonomy
  type: has_many
  via: taxonomies
- from: Taxonomy
  to: PostType
  type: has_many
  via: types
traversal_shortcuts:
  _embed: >-
    Appending _embed to a collection or object request inlines author, featured media and terms
    under _embedded, collapsing the has_one/has_many hops above into a single request.
  _fields: >-
    Appending _fields=id,slug,title,link,acf trims the payload substantially — the default
    representation of every object carries a large yoast_head HTML blob plus yoast_head_json.
subway: null