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

# import_prompts

> Import tracked questions into the brand's question bank (additive — never modifies or removes existing questions).

Import tracked questions into the brand's question bank (additive — never modifies or removes existing questions). Each item needs the question text plus one or more personas from plan\_question\_bank; journeyStage is optional (auto-classified when omitted; one of: awareness\_discovery, consideration\_evaluation, purchase\_intent, support\_usage).

Give each question every persona a real buyer of that type would plausibly ask it (personaIds or personaNames arrays), and only those. A single personaId/personaName still works. Don't spray every persona onto every question.

Every (question, persona) is tracked and measured separately and consumes one plan slot. A question attached to 3 personas is 3 tracked questions, and the same question can score differently for each.

Server rules:

* Net-new (question, persona) rows must fit remaining headroom (duplicates cost nothing). Over-cap imports are rejected with the exact headroom.
* A (question, persona) pair already tracked is skipped as duplicate — never overwritten.
* Max 50 items per call (each item may still fan out to several personas).

ALWAYS run \{ dryRun: true } first and show the user the result — then import for real only after their explicit confirmation. Imported questions enter the next scheduled scan automatically.

## Inputs

| Field | Type | Required | Default | Description |
| - | - | - | - | - |
| `prompts` | `object[]` | yes | — | The questions to import. |
| `dryRun` | `boolean` | no | — | true = validate and report what WOULD happen without writing. Always dry-run first. |


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.