> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bland.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Get Agent Memory Schema

> Read what the agent remembers about each contact between calls.

### Overview

Returns the memory entity schema the agent runs with today: the kinds of things it tracks per contact, such as a booking or an order, and the fields it fills in for each one.

The schema comes from one of two places, and `source` tells you which. When the newest dev version carries `settings.memorySchema`, that is returned and `source` is `settings`. Otherwise the last schema the runtime inferred and cached is returned with `source` set to `cached_version`. An agent with neither returns an empty list and `source` of `null`, which means nothing is known yet rather than that the agent remembers nothing.

To produce a schema for an agent that has none, use [Infer Agent Memory Schema](/api-v2/post/agents-id-memory-schema-infer).

### Headers

<ParamField header="authorization" type="string" required>
  Your API key for authentication.
</ParamField>

### Path Parameters

<ParamField path="agent_id" type="string" required>
  The agent's unique identifier.
</ParamField>

### Response

<ResponseField name="data.entity_schemas" type="array">
  The entity types the agent tracks per contact. Empty when nothing is known yet.

  <Expandable title="entity schema object">
    <ResponseField name="entity_type" type="string">
      Name of the entity, for example `trip`, `order`, or `quote`.
    </ResponseField>

    <ResponseField name="description" type="string">
      What this entity represents.
    </ResponseField>

    <ResponseField name="identifier_field" type="string">
      The field used to tell one instance from another, for example `confirmation_number`.
    </ResponseField>

    <ResponseField name="fields" type="array">
      Field definitions. Each has `name`, `description`, and `type`, plus optional `enum_values`, `required`, `array_item_type`, `array_item_schema`, `type_conversion`, and `conversational_format`.
    </ResponseField>

    <ResponseField name="status_field" type="string">
      Which field tracks the entity's lifecycle. Present only when one is set.
    </ResponseField>

    <ResponseField name="outcomes" type="array">
      Possible end states for this entity. Present only when any are set.
    </ResponseField>

    <ResponseField name="metadata" type="object">
      Extra context about the entity type. Present only when set.
    </ResponseField>
  </Expandable>
</ResponseField>

<ResponseField name="data.source" type="string | null">
  Where the schema came from: `settings` when it is saved on the agent's newest dev version, `cached_version` when it was inferred by the runtime, or `null` when `entity_schemas` is empty.
</ResponseField>

<ResponseField name="errors" type="null | array">
  `null` on success, or a list of error objects if the request failed.
</ResponseField>

<ResponseExample>
  ```json Response theme={null}
  {
    "data": {
      "entity_schemas": [
        {
          "entity_type": "appointment",
          "description": "A dental appointment the caller books, moves, or cancels.",
          "identifier_field": "confirmation_number",
          "fields": [
            {
              "name": "confirmation_number",
              "description": "The booking reference given to the caller.",
              "type": "string",
              "required": true
            },
            {
              "name": "starts_at",
              "description": "When the appointment starts.",
              "type": "string",
              "type_conversion": {
                "natural_language_examples": ["tomorrow at 2", "next Tuesday morning"],
                "target_format": "ISO 8601 date-time"
              }
            }
          ],
          "status_field": "state",
          "outcomes": ["booked", "rescheduled", "cancelled"]
        }
      ],
      "source": "settings"
    },
    "errors": null
  }
  ```

  ```json Nothing Known Yet theme={null}
  {
    "data": {
      "entity_schemas": [],
      "source": null
    },
    "errors": null
  }
  ```

  ```json Not Found theme={null}
  {
    "data": null,
    "errors": [
      {
        "error": "NOT_FOUND",
        "message": "Agent not found"
      }
    ]
  }
  ```
</ResponseExample>

***

Docs for agents: [llms.txt](/llms.txt)