generated: '2026-07-21'
method: derived
source: openapi/vijil-openapi-original.json
note: >-
  Entity-relationship graph derived from the Vijil Console OpenAPI: schema
  properties and *_id reference fields. Team is the top-level tenant; nearly
  every resource is team-scoped. Identifiers are UUIDs.
entities:
  - name: Team
    description: Top-level tenant/workspace that owns all other resources.
    id: team_id
  - name: User
    description: An account; joins Teams via Membership.
    id: user_id / external_user_id
  - name: Membership
    description: Join between User and Team, carrying a TeamRole.
    id: membership_id
  - name: Invitation
    description: Pending invite of a user to a Team.
  - name: Agent
    description: A registered AI agent/model configuration under evaluation and protection.
    id: id (agent_id)
  - name: ApiKey
    description: Stored model-hub credential (OpenAI/Together/custom) used to query the agent's model.
    id: api_key_id
  - name: Evaluation
    description: A Diamond trust evaluation run against an Agent (async job).
    id: evaluation_id
  - name: EvaluationSummary
    description: Rolled-up scores for an Evaluation.
  - name: Harness
    description: A standard test suite (safety, security, reliability, privacy, toxicity, ethics).
  - name: CustomHarness
    description: A team-authored evaluation suite built from Personas/Policies.
    id: harness id / workflow_id
  - name: Persona
    description: A user profile used to drive evaluations/red-team.
  - name: Policy
    description: Organizational rules an Agent must follow during evaluation.
  - name: PolicyRule
    description: An individual rule within a Policy.
    id: rule_id
  - name: DomeConfig
    description: A runtime Guardrail (Dome) configuration, bindable to an Agent.
    id: dome_config_id
  - name: Detection
    description: A Dome detection request/result.
    id: detector_id
  - name: Campaign
    description: A Red Team adversarial campaign against an Agent (async job).
    id: campaign_id
  - name: Genome
    description: Darwin evolution genome extracted from / applied to an Agent.
    id: genome_id
  - name: Proposal
    description: A Darwin mutation proposal for an Agent's genome.
  - name: DemographicDimension
    description: Named demographic dimension + values used in evaluation simulation.
relationships:
  - {from: Membership, to: Team, type: belongs_to, via: team_id}
  - {from: Membership, to: User, type: belongs_to, via: user_id}
  - {from: Invitation, to: Team, type: belongs_to, via: team_id}
  - {from: Agent, to: Team, type: belongs_to, via: team_id}
  - {from: Agent, to: DomeConfig, type: has_one, via: dome_config_id}
  - {from: Agent, to: ApiKey, type: has_one, via: api_key_id}
  - {from: Agent, to: Evaluation, type: has_many, via: evaluation_id (AgentLatestEvaluation)}
  - {from: Evaluation, to: Agent, type: belongs_to, via: agent_id}
  - {from: Evaluation, to: Team, type: belongs_to, via: team_id}
  - {from: EvaluationSummary, to: Agent, type: belongs_to, via: agent_id}
  - {from: EvaluationSummary, to: Team, type: belongs_to, via: team_id}
  - {from: CustomHarness, to: Agent, type: belongs_to, via: agent_id}
  - {from: CustomHarness, to: Team, type: belongs_to, via: team_id}
  - {from: CustomHarness, to: Persona, type: has_many, via: persona}
  - {from: Persona, to: Team, type: belongs_to, via: team_id}
  - {from: Policy, to: Team, type: belongs_to, via: team_id}
  - {from: PolicyRule, to: Policy, type: belongs_to, via: policy_id}
  - {from: DomeConfig, to: Agent, type: belongs_to, via: agent_id}
  - {from: Detection, to: DomeConfig, type: uses, via: detector_id}
  - {from: Campaign, to: Agent, type: belongs_to, via: agent_id}
  - {from: Campaign, to: Team, type: belongs_to, via: team_id}
  - {from: Proposal, to: Agent, type: belongs_to, via: agent_id}
  - {from: Proposal, to: Genome, type: belongs_to, via: genome_id}
  - {from: Proposal, to: Team, type: belongs_to, via: team_id}
  - {from: Genome, to: Agent, type: belongs_to, via: agent_id}