Skip to main content

​
Overview

Core types are predefined GraphQL types provided by Exulu IMP for managing platform resources. These types are always available in the schema, regardless of your implementation.

​
Agent

​
agent

AI agent configurations and instances. 
type agent {
  id: ID!
  name: String!
  description: String
  backend: String!
  type: String!
  createdAt: Date!
  updatedAt: Date!

  # Backend-specific fields
  provider: String
  providerName: String
  modelName: String
  systemInstructions: String
  authenticationInformation: String

  # Configuration
  rateLimit: RateLimiterRule
  streaming: Boolean
  capabilities: AgentCapabilities
  maxContextLength: Int

  # Workflows
  workflows: AgentWorkflows

  # Access control (if RBAC enabled)
  RBAC: RBACData

  # Routing
  slug: String
}
Example query: 
query {
  agentById(id: "agent-123") {
    id
    name
    provider
    modelName
    capabilities {
      text
      images
      files
    }
    workflows {
      enabled
      queue {
        name
      }
    }
  }
}

​
AgentCapabilities

Defines what content types the agent can process. 
type AgentCapabilities {
  text: Boolean
  images: [String]    # e.g., [".png", ".jpg"]
  files: [String]     # e.g., [".pdf", ".docx"]
  audio: [String]
  video: [String]
}

​
RateLimiterRule

Rate limiting configuration for agent requests. 
type RateLimiterRule {
  name: String
  rate_limit: RateLimiterRuleRateLimit
}

type RateLimiterRuleRateLimit {
  time: Int    # Time window in seconds
  limit: Int   # Max requests per time window
}

​
Session types

​
agent_sessions

Conversation sessions between users and agents. 
type agent_sessions {
  id: ID!
  agent: ID!
  user: ID
  name: String
  createdAt: Date!
  updatedAt: Date!
  RBAC: RBACData
}
Example query: 
query {
  agent_sessionsPagination(
    limit: 20
    filters: [{ agent: { eq: "agent-123" } }]
  ) {
    items {
      id
      name
      createdAt
    }
  }
}

​
agent_messages

Individual messages within a session. 
type agent_messages {
  id: ID!
  session: ID!
  role: String!       # "user" | "assistant" | "system"
  content: String!
  createdAt: Date!
  updatedAt: Date!
}

​
User management types

​
user

User accounts for authentication and access control. 
type user {
  id: Int!
  name: String
  email: String!
  emailVerified: Date
  image: String
  password: String
  type: String        # "api" | "user"
  apikey: String
  role: ID
  super_admin: Boolean
  createdAt: Date!
  updatedAt: Date!
}
Example query: 
query {
  usersPagination(
    filters: [{ type: { eq: "user" } }]
  ) {
    items {
      id
      name
      email
      role
      super_admin
    }
  }
}
The password and apikey fields are write-only for security. They are hashed with bcrypt and cannot be retrieved.

​
role

Role definitions for RBAC. 
type role {
  id: ID!
  name: String!
  agents: String      # "read" | "write"
  evals: String
  workflows: String
  variables: String
  users: String
  api: String
  createdAt: Date!
  updatedAt: Date!
}
Example mutation: 
mutation {
  rolesCreateOne(
    input: {
      name: "Developer"
      agents: "write"
      evals: "write"
      workflows: "read"
      variables: "read"
      users: "read"
    }
  ) {
    item {
      id
      name
    }
  }
}

​
Evaluation types

​
test_cases

Test cases for evaluating agent performance. 
type test_cases {
  id: ID!
  name: String!
  description: String
  inputs: JSON!                           # UIMessage[]
  expected_output: String!
  expected_tools: JSON                    # string[]
  expected_knowledge_sources: JSON        # string[]
  expected_agent_tools: JSON              # string[]
  createdAt: Date!
  updatedAt: Date!
  RBAC: RBACData
}
Example: 
mutation {
  test_casesCreateOne(
    input: {
      name: "Weather Query"
      description: "User asks about weather"
      inputs: [
        { role: "user", content: "What's the weather like?" }
      ]
      expected_output: "Based on current data, it's 68°F and sunny."
      expected_tools: ["get_weather"]
    }
  ) {
    item {
      id
      name
    }
  }
}

​
eval_sets

Collections of test cases for batch evaluation. 
type eval_sets {
  id: ID!
  name: String!
  description: String
  test_case_ids: JSON       # string[]
  createdAt: Date!
  updatedAt: Date!
  RBAC: RBACData
}

​
eval_runs

Evaluation execution records. 
type eval_runs {
  id: ID!
  name: String!
  description: String
  agent_id: ID!
  test_case_ids: JSON           # string[]
  eval_functions: JSON          # string[]
  config: JSON
  scoring_method: String
  pass_threshold: Float
  timeout_in_seconds: Int
  createdAt: Date!
  updatedAt: Date!
  RBAC: RBACData
}

​
Workflow types

​
workflow_templates

Workflow definitions with agent execution steps. 
type workflow_templates {
  id: ID!
  name: String!
  description: String
  agent: ID!
  steps_json: JSON
  variables: [String]     # Extracted from steps_json
  createdAt: Date!
  updatedAt: Date!
  RBAC: RBACData
}
Example query: 
query {
  workflow_templateById(id: "workflow-123") {
    id
    name
    agent
    variables
  }
}

​
Configuration types

​
variables

Encrypted configuration values (alternative to .env files). 
type variables {
  id: ID!
  name: String!
  value: String!
  encrypted: Boolean!
  description: String
  createdAt: Date!
  updatedAt: Date!
}
Values are encrypted with AES using NEXTAUTH_SECRET. Use ExuluVariables.get() to retrieve decrypted values.

​
projects

Project organization for grouping resources. 
type projects {
  id: ID!
  name: String!
  description: String
  createdAt: Date!
  updatedAt: Date!
}

​
Prompt management types

​
prompt_library

Saved prompts for reuse across agents. 
type prompt_library {
  id: ID!
  name: String!
  content: String!
  tags: JSON          # string[]
  category: String
  description: String
  createdAt: Date!
  updatedAt: Date!
  RBAC: RBACData
}
Example query: 
query {
  prompt_libraryPagination(
    filters: [
      { tags: { contains: ["support", "greeting"] } }
    ]
  ) {
    items {
      id
      name
      content
      tags
    }
  }
}

​
prompt_favorites

User’s favorited prompts. 
type prompt_favorites {
  id: ID!
  user: ID!
  prompt: ID!
  createdAt: Date!
  updatedAt: Date!
}

​
Analytics types

​
statistics

Usage statistics and analytics. 
type statistics {
  id: ID!
  name: String!
  label: String!
  type: String!
  trigger: String!
  count: Int!
  user: ID
  role: ID
  createdAt: Date!
  updatedAt: Date!
}
Aggregate statistics: 
query {
  statisticsStatistics(
    groupBy: "label"
    filters: [
      { type: { eq: "agent_run" } }
    ]
    limit: 10
  ) {
    group
    count
  }
}

​
Job management types

​
Job

Background job information from BullMQ. 
type Job {
  id: String!
  name: String!
  returnvalue: JSON
  stacktrace: [String]
  finishedOn: Date
  processedOn: Date
  attemptsMade: Int
  failedReason: String
  state: String!
  data: JSON
  timestamp: Date
}
Query jobs: 
query {
  jobs(
    queue: eval_runs
    statusses: [active, waiting, failed]
    page: 1
    limit: 50
  ) {
    items {
      id
      name
      state
      failedReason
    }
  }
}

​
RBAC

​
RBACData

Access control information for resources. 
type RBACData {
  type: String!       # "public" | "private" | "users" | "roles"
  users: [RBACUser!]
  roles: [RBACRole!]
}

type RBACUser {
  id: ID!
  rights: String!     # "read" | "write"
}

type RBACRole {
  id: ID!
  rights: String!     # "read" | "write"
}

​
RBACInput

Input type for setting RBAC permissions. 
input RBACInput {
  users: [RBACUserInput!]
  roles: [RBACRoleInput!]
}

input RBACUserInput {
  id: ID!
  rights: String!     # "read" | "write"
}

input RBACRoleInput {
  id: ID!
  rights: String!     # "read" | "write"
}

​
Next steps

Dynamic types

Learn about context-generated types

Queries

Explore all query operations

Mutations

Explore all mutation operations

Introduction

Back to API overview