> ## Documentation Index
> Fetch the complete documentation index at: https://resources.athenaintel.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Object Types
> Define the entities that matter to your organization with typed properties and validation rules.
An object type is a schema that describes a category of real-world entity — a company, a deal, a person, a product. It defines what properties each instance will have, what types those properties are, and which constraints apply.
## Creating an Object Type
From the [Ontology Builder](/docs/ontology/ontology-builder), click **New Object Type** in the top-right corner, or double-click anywhere on the canvas.
Enter a **Name** for the object type (e.g., "Deal", "Client", "Candidate"). Optionally add a **Description** explaining what this type represents and an **Icon URL** for visual identification.
Define the properties (fields) for this object type. Each property needs a name and type. You can add constraints like required, default values, and min/max limits.
Select one of the required text properties as the **Title Property**. This is the display name shown in search results, tabs, and mentions.
Click **Save** (or **Create** for new types). The object type is immediately available for agents and the Ontology Builder graph.
## Templates
To get started quickly, Athena provides built-in templates for common entity types. Click any template to pre-populate the property list with a sensible starting schema.
Contact or individual. Properties: first name, last name, email, phone, title, company, LinkedIn URL, status.
Organization or business. Properties: name, legal name, domain, industry, founded year, employee count, revenue, type.
Sales opportunity. Properties: deal name, account, contact, value, probability, close date, owner, pipeline stage.
Client account. Properties: company name, primary contact, industry, website, revenue, employee count, tier.
Hiring candidate. Properties: full name, email, position applied, years experience, expected salary, pipeline stage.
Physical location. Properties: name, address, city, state, postal code, country, coordinates, location type.
Templates are a starting point — you can add, remove, or modify any property after loading a template.
## Property Types
Every property has a type that determines what values it can hold and how it's validated.
Free-form string values. Use for names, emails, URLs, descriptions, and notes.
**Optional constraints:**
* **Min Length** / **Max Length** — restrict the character count
* **Pattern** — a regular expression the value must match (e.g., email format)
```
firm_name: Text (required)
email: Text
notes: Text (max length: 2000)
```
Decimal numeric values. Use for revenue, prices, coordinates, percentages.
**Optional constraints:**
* **Minimum** / **Maximum** — restrict the value range
```
annual_revenue: Number (min: 0)
latitude: Number (min: -90, max: 90)
deal_value: Number (min: 0)
```
Whole number values. Use for counts, years, quantities.
**Optional constraints:**
* **Minimum** / **Maximum** — restrict the value range
```
employee_count: Integer (min: 1)
founded_year: Integer (min: 1800, max: 2030)
quantity: Integer (min: 0)
```
True or false values. Use for flags and binary states.
```
is_active: Boolean
is_verified: Boolean
```
Calendar date values. Use for deadlines, start dates, timestamps.
```
close_date: Date
contract_start: Date
date_of_birth: Date
```
A fixed set of allowed values. Use for statuses, categories, tiers, and pipeline stages.
**Configuration:**
* **Enum Values** — the list of allowed options
* **Default Value** — the initial value for new instances
```
stage: Enum [prospecting, qualification, proposal, negotiation, closed_won, closed_lost]
tier: Enum [enterprise, mid_market, smb, startup] (default: mid_market)
```
## Title Property
Every object type must have a **title property** — a required text property that serves as the display name for each instance. The title property is shown:
* In **search results** when querying ontology data
* In **tab headers** when viewing an instance in Spaces
* In **mentions** when referencing an instance in conversations
* In **relationship displays** when showing linked objects
Only properties that are both **required** and **text type** are eligible to be the title property. If you don't have one yet, create a required text property first.
## Property Builder
The property builder provides a drag-and-drop interface for defining and ordering properties:
* **Reorder** properties by dragging the handle on the left
* **Expand** any property to configure its type, description, constraints, and required/default settings
* **Delete** a property by clicking the remove icon
* **Add** new properties with the "Add Property" button at the bottom
Properties are automatically synced to a JSON Schema representation, visible in the right panel of the editor. You can also edit the JSON Schema directly — changes are reflected back in the property builder.
## JSON Schema
Under the hood, each object type's properties are stored as a [JSON Schema](https://json-schema.org/). The editor's right panel shows a live preview of the generated schema:
```json theme={null}
{
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"title": "Deal",
"properties": {
"deal_name": {
"type": "string",
"description": "Deal or opportunity name"
},
"deal_value": {
"type": "number",
"description": "Total deal value in USD",
"minimum": 0
},
"stage": {
"type": "string",
"enum": ["prospecting", "qualification", "proposal", "negotiation", "closed_won", "closed_lost"],
"description": "Pipeline stage"
}
},
"required": ["deal_name", "stage"]
}
```
You can switch between the visual property builder and the raw JSON editor at any time. Edits in one are reflected in the other.
### Importing a Schema
If you already have a JSON Schema from another tool or specification, click the **Import JSON Schema** button in the template bar. Paste your schema and click **Import** — the properties will be populated automatically. This is useful for bootstrapping object types from existing data models.
## Editing and Deleting
* To **edit** an existing object type, click on its node in the Ontology Builder graph, or double-click it to open the editor directly.
* To **delete** an object type, use the delete action from the node's context menu. Deleting an object type also permanently deletes all its instances and any relationships that reference it.
Deleting an object type is permanent. All instances and connected relationships will be removed and cannot be recovered.