generated: '2026-08-25'
method: derived
source: openapi/ (9 documents) + live anonymous responses and /wp/v2/types + /wp/v2/taxonomies read 2026-08-25
note: >-
  Entity-relationship graph of the MBrace Therapeutics content API, derived from the registered post
  types and taxonomies the deployment reports about itself and from the id-reference fields actually
  present in live response bodies. The interesting part of this model is not the WordPress core
  types — it is the five custom post types MBrace registered, which make the company's governance
  and cap-table disclosures queryable as structured records.
identifiers:
  scheme: integer
  detail: >-
    Every object is keyed by a site-scoped integer `id`. Identifiers are NOT prefixed or typed, so
    an id is only meaningful together with the collection it came from — id 104 is a media
    attachment, id 2787 is an investor. /wp/v2/search returns {id, type, subtype} precisely so a
    caller can pick the right collection to dereference against.
  slugs: >-
    Every object also carries a `slug`, unique within its type, and every collection accepts
    ?slug= as a lookup. Slugs are the stable human-readable key; ids are the stable machine key.
entities:
- name: Post
  collection: /wp/v2/posts
  count_observed: 8
  description: A company news item — press release, conference presentation or publication.
  key_fields: [id, slug, title, content, excerpt, date, modified, link, author, featured_media, categories, tags]
  relationships:
  - type: belongs_to
    target: User
    via: author
  - type: has_one
    target: MediaItem
    via: featured_media
  - type: has_many
    target: Category
    via: categories
  - type: has_many
    target: Tag
    via: tags
    note: Always empty — the post_tag taxonomy carries zero terms on this deployment.
- name: Page
  collection: /wp/v2/pages
  count_observed: 12
  description: >-
    A corporate site page — About Us, Pipeline, Discovery Platform, the MBRC-101 and MBRC-201
    programme pages, Expanded Access Statement, Contact, Terms of Use, Privacy Policy.
  key_fields: [id, slug, title, content, excerpt, parent, menu_order, date, modified, link, author, featured_media]
  relationships:
  - type: belongs_to
    target: Page
    via: parent
    note: Self-referential hierarchy.
  - type: belongs_to
    target: User
    via: author
  - type: has_one
    target: MediaItem
    via: featured_media
- name: MediaItem
  collection: /wp/v2/media
  count_observed: 224
  description: An uploaded attachment — logos, leadership headshots, document assets.
  key_fields: [id, slug, title, source_url, mime_type, media_type, media_details, filename, filesize, alt_text, caption, post, author]
  relationships:
  - type: belongs_to
    target: Post
    via: post
    note: The object the attachment was uploaded to. Polymorphic — may reference a page or a custom post type.
  - type: belongs_to
    target: User
    via: author
- name: BoardMember
  collection: /wp/v2/board-member
  count_observed: 8
  description: A member of the MBrace Therapeutics board of directors.
  custom_post_type: true
  key_fields: [id, slug, title, content, excerpt, parent, featured_media, categories]
  detail: >-
    `title.rendered` carries the name with post-nominals (observed: "Fabian Zohren, MD, PhD");
    `content.rendered` carries the biography HTML; `featured_media` resolves to the headshot.
  relationships:
  - type: has_one
    target: MediaItem
    via: featured_media
  - type: has_many
    target: Category
    via: categories
- name: ExecutiveCommitteeMember
  collection: /wp/v2/executive-committee
  count_observed: 8
  description: A company officer on the executive committee.
  custom_post_type: true
  key_fields: [id, slug, title, content, excerpt, featured_media, categories]
  relationships:
  - type: has_one
    target: MediaItem
    via: featured_media
  - type: has_many
    target: Category
    via: categories
- name: Founder
  collection: /wp/v2/founders
  count_observed: 2
  description: >-
    A company co-founder. Observed at harvest time: Renata Pasqualini, Ph.D. and Isan Chen, MD.
  custom_post_type: true
  key_fields: [id, slug, title, content, excerpt, parent, featured_media, categories]
  relationships:
  - type: has_one
    target: MediaItem
    via: featured_media
  - type: has_many
    target: Category
    via: categories
- name: Investor
  collection: /wp/v2/investor
  count_observed: 5
  description: >-
    A named institutional backer. Observed at harvest time: Blue Owl Capital, Avidity Partners, TPG,
    Venrock and Alta Partners.
  custom_post_type: true
  key_fields: [id, slug, title, content, featured_media]
  detail: >-
    The only one of the five custom types NOT attached to the category taxonomy, and the only one
    with no `excerpt` field in its response. `featured_media` resolves to the investor's logo.
  relationships:
  - type: has_one
    target: MediaItem
    via: featured_media
- name: ScientificBoardMember
  collection: /wp/v2/scientific-board
  count_observed: 5
  description: A member of the scientific advisory board — clinical and translational oncology advisers.
  custom_post_type: true
  key_fields: [id, slug, title, content, excerpt, parent, featured_media, meta]
  relationships:
  - type: has_one
    target: MediaItem
    via: featured_media
- name: Category
  collection: /wp/v2/categories
  count_observed: 4
  description: A category term.
  key_fields: [id, slug, name, description, parent, count, taxonomy, link]
  detail: >-
    Unusually, this taxonomy is registered against FOUR post types — post, board-member,
    executive-committee and founders — so it is the shared classifier spanning the news archive and
    the people collections, not just a blog taxonomy.
  relationships:
  - type: belongs_to
    target: Category
    via: parent
    note: Self-referential hierarchy.
- name: Tag
  collection: /wp/v2/tags
  count_observed: 0
  description: A post_tag term. Registered and reachable, but the collection is empty.
- name: User
  collection: /wp/v2/users
  count_observed: 3
  description: >-
    A content author. Anonymously readable on this deployment — display fields only (name, slug,
    description, link, avatar_urls). No email addresses or capabilities are exposed.
  key_fields: [id, slug, name, description, link, avatar_urls]
- name: SearchResult
  collection: /wp/v2/search
  count_observed: 76
  description: >-
    A lightweight reference, not a stored entity. {id, title, url, type, subtype} — the resolver
    between a query string and the typed collections above.
  relationships:
  - type: polymorphic_reference
    target: any
    via: id + subtype
graph_notes:
- >-
  The five custom post types are the reason this profile exists. A clinical-stage biotech that
  publishes its board, executive committee, founders, named investors and scientific advisory board
  as five separate JSON collections is emitting a machine-readable governance and cap-table
  disclosure, whether or not it meant to.
- >-
  The graph is shallow and read-only. There are no join entities, no write paths, no state machines
  and no lifecycle transitions — every relationship is a single-hop id reference resolved with a
  second GET.
- >-
  The `acf` (Advanced Custom Fields) bridge is active on every entity above but returned an EMPTY
  array on all of them on 2026-08-25. That is the model's biggest gap: whatever structured fields
  the site editors use to render titles, roles and affiliations are NOT exposed through the API, so
  a consumer must parse `title.rendered` and `content.rendered` HTML to recover them.
- >-
  No entity carries an MBRC programme identifier. The pipeline candidates (MBRC-101, MBRC-201,
  MBRC-301, MBRC-401) exist only as page content, not as a modelled entity — there is no
  /wp/v2/pipeline or /wp/v2/program collection. An agent wanting pipeline data must read the Pages
  API and parse HTML.
cross_links:
  conventions: conventions/mbrace-therapeutics-conventions.yml
  errors: errors/mbrace-therapeutics-problem-types.yml