generated: '2026-09-19'
method: derived
source: openapi/yuens-me-openapi.yml (https://agent.yuens.me/openapi.json) response schemas plus the README's endpoint descriptions; no live profile data was copied.
description: >-
  A single-candidate data model: one Profile owns the structured snapshot, Projects and Observations are the
  evidence layer, and two computed results (QueryAnswer, MatchResult) cite back into it by slug. There are no
  ids in the public contract apart from Observation ids and Project / Publication slugs; the schemas are
  inline (components.schemas is empty), so relationships below are read from field names.
entities:
  - name: Profile
    operation: getProfile
    fields: ['contact', 'summary', 'tagline', 'skills[]', 'employment[]', 'education[]', 'projects[]', 'publications[]', 'availability']
    note: contact is an object of the person's published contact details; treat as personal data and never cache it beyond the request.
  - name: Availability
    operation: getAvailability
    fields: ['seeking', 'status(open|actively-looking|not-looking)', 'preferred_roles[]', 'remote']
  - name: Project
    operation: listProjects
    key: slug
    fields: ['name', 'slug', 'description', 'role', 'tech[]', 'status', 'started', 'url', 'repo', 'cover']
  - name: Observation
    operation: listObservations
    key: id
    fields: ['id', 'type(observation|idea|task|reference)', 'authored(boolean)', 'topics', 'date', 'content']
    url_pattern: https://agent.yuens.me/observations/{id}
    envelope: '{count, total, truncated, scope, types, authored?, note, observations[]}'
  - name: Publication
    key: slug
    fields: [slug, title, platform, canonical_url, date]
    note: Appears inline in Profile.publications[] and in QueryAnswer.publications[]; no standalone endpoint.
  - name: QueryAnswer
    operation: queryProfile
    computed: true
    fields: ['answer', 'confidence(high|medium|low)', 'sources[]', 'project_slugs[]', 'publications[]', 'follow_up_suggestions[]', 'action_intent{tool}?', 'fit_question']
  - name: MatchResult
    operation: matchJob
    computed: true
    fields: ['fit_score(0-1)', 'matched[]', 'gaps[]', 'verdict', 'recommended_action(apply|apply-with-tailoring|pass)', 'scoring{required_qualities[], scored_qualities[]}']
relationships:
  - from: Profile
    to: Project
    type: has_many
    via: projects[]
  - from: Profile
    to: Publication
    type: has_many
    via: publications[]
  - from: Profile
    to: Availability
    type: has_one
    via: availability
  - from: QueryAnswer
    to: Project
    type: belongs_to
    via: project_slugs[] -> Project.slug
  - from: QueryAnswer
    to: Publication
    type: belongs_to
    via: publications[].slug -> Publication.slug
  - from: QueryAnswer
    to: Observation
    type: references
    via: 'sources[] entries of the form observations:"<excerpt>" (prose citation, not an id)'
  - from: MatchResult
    to: Profile
    type: belongs_to
    via: scored_qualities[].evidence_grade (verified = backed by a dated project or employment record)
identifiers:
  slug: lowercase kebab-case project / publication key (e.g. resume-agent, brew-guide)
  observation_id: opaque stable id, individually addressable at /observations/{id}