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

# Panels

> 1 MCP tool for panel provisioning.

Panel provisioning for recruiting participants from a structured panel. 1 tool.

| Tool           | Description                                                           |
| -------------- | --------------------------------------------------------------------- |
| `create_panel` | Provision a panel survey for a **Panel-type study** (`is_panel=true`) |

<Note>
  `create_panel` is valid only for **Panel studies** (`is_panel=true`). A **BYOP** ("bring your own participant", `is_panel=false`) study cannot field a panel — it recruits its own participants via `create_invite`. The tool reads the study first and refuses BYOP studies. See [Studies → Panel vs BYOP](/mcp-server/tools/studies).
</Note>

### Dry run

Pass `dry_run: true` to `create_panel` to preview the full cost **without** provisioning the panel, placing a hold, or charging — mirroring the `ask_humans` dry-run. The estimate covers both charges:

* **Credits** (per-response cost): `per_response_cost`, `estimated_cost`, `available_balance`, and `sufficient_credits`.
* **Card reward-hold**: `hold_amount_usd` (= target × per-participant reward, captured at settlement) plus `hold_required` — the authorization placed on the card at launch, separate from the credit cost. It is `0` for enterprise plans or wallets with no saved card.

## How User Intuition handles this

`create_panel` provisions a panel recruit when you need respondents beyond your own customer base. See [`/platform/agentic-research/`](https://www.userintuition.ai/platform/agentic-research/) for the end-to-end agent workflow.