> ## 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 Variable Catalog

> List the agent variables a disposition can re-extract.

### Overview

<Note>
  Dispositions are enabled per organization. If your organization does not have access, these endpoints return `404`.
</Note>

Lists the variables the agent extracts during calls that a disposition value can re-extract from the corrected post-call transcript, read from the agent's most recently saved version. Only `string`, `number`, and `boolean` variables are included; `json` variables, and variables defined with conflicting prompts or types in different scenarios, are omitted. Entries are sorted by `key`, and the list is empty if the agent has no saved version.

### 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. A malformed id returns `400` with the message `agentId must be a valid UUID`.
</ParamField>

### Response

<ResponseField name="data" type="array">
  Array of variable catalog entries.

  <Expandable title="variable entry">
    <ResponseField name="key" type="string">
      The variable name as defined on the agent. Use it as `variableKey` in a disposition value's `variable_extraction` source.
    </ResponseField>

    <ResponseField name="prompt" type="string">
      The extraction prompt configured on the agent, trimmed. Empty string when the variable has no prompt.
    </ResponseField>

    <ResponseField name="accurateSpelling" type="boolean">
      `true` when the variable is configured for exact character-by-character spelling.
    </ResponseField>

    <ResponseField name="outputSchema" type="object">
      The schema a disposition value should use for this variable. `kind` mirrors the variable's type (`string`, `number`, or `boolean`). `description` is the prompt, with the sentence `Preserve exact character-by-character spelling.` appended when `accurateSpelling` is `true`; it is omitted when both are empty.
    </ResponseField>
  </Expandable>
</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": [
      {
        "key": "callback_number",
        "prompt": "The phone number the caller wants us to call back.",
        "accurateSpelling": true,
        "outputSchema": {
          "kind": "string",
          "description": "The phone number the caller wants us to call back. Preserve exact character-by-character spelling."
        }
      },
      {
        "key": "party_size",
        "prompt": "How many people the reservation is for.",
        "accurateSpelling": false,
        "outputSchema": {
          "kind": "number",
          "description": "How many people the reservation is for."
        }
      },
      {
        "key": "wants_followup",
        "prompt": "Whether the caller asked for a follow-up call.",
        "accurateSpelling": false,
        "outputSchema": {
          "kind": "boolean",
          "description": "Whether the caller asked for a follow-up call."
        }
      }
    ],
    "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)