> For the complete documentation index, see [llms.txt](https://help.metaforms.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.metaforms.ai/tabulations/using-the-dialing-agent-for-quick-table-setup.md).

# Using the Tabulation Agent

{% embed url="<https://www.youtube.com/watch?v=zUpceW8rwiw>" %}

The Tabulation Agent makes changes to your tables from plain-language requests. It can build and edit tables, banners and statistics, explain results and fix issues. You need tabulation edit access to use it.

### Asking the agent

Step 1: Click the agent icon at the top of the right rail.

Step 2: Type your request. Press Enter to send, or Shift+Enter for a new line.

Step 3: Read the reply. The agent says what it changed and the deck moves to a new plan version.

![An agent reply after a change](https://1402057010-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FgtHd8o9ldznZ1hKHiZvc%2Fuploads%2Fgit-blob-2ceac0c4e40d195d688a793edd9951d978926df2%2Ftabulations-agent-chat-example.png?alt=media)

Click **+** to start a new chat. To ask about one table, choose **Ask the agent about this table** from the table menu. This adds the table's ID and title to your message, ready for you to finish.

### What you can ask

Click the checklist icon in the right rail to see example requests.

![What I can help with](https://1402057010-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FgtHd8o9ldznZ1hKHiZvc%2Fuploads%2Fgit-blob-5462928201dc5cf31023e7c8c71624b53a36d0ec%2Ftabulations-agent-capabilities.png?alt=media)

If you have no tab spec, the agent can draft one from your data. Ask, for example, "Draft a tab spec from this data". It shows an approval card first, and after you approve, the draft goes through [mapping review](https://help.metaforms.ai/tabulations/reviewing-the-tab-spec-mapping.md) like an uploaded spec. The status bar marks the file **AI generated**.

### Changes the agent applies, and changes you approve

The agent applies these directly: nets, bases, filters, weights, significance levels, statistics, reordering, labels and row visibility.

For these it shows an approval card first: defining, editing, nesting or removing banners, changing factors or scales, changing table visibility, adding tables, applying a spec mapping and generating a tab spec.

The card is headed **Approval needed**. **What this changes** says what will happen, and **Needs your judgment** lists anything the agent could not decide for you.

![An approval card](https://1402057010-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FgtHd8o9ldznZ1hKHiZvc%2Fuploads%2Fgit-blob-3948592a7d2714edb2e1a82cb9f0626564064c6c%2Ftabulations-agent-approval-card.png?alt=media)

You can **Add a note (optional)**, then click **Approve** or **Reject**. Nothing changes until you approve.

### What the agent will not do

The agent does not work on trackers or multi-wave studies, or on migrating older projects. When a request is out of scope, it shows an **Out of scope** card with the reason and what you can do instead. It has no access to respondent-level data.

![An out of scope card](https://1402057010-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FgtHd8o9ldznZ1hKHiZvc%2Fuploads%2Fgit-blob-e4976281b34ad21ce967e437cdc45149a643eb48%2Ftabulations-agent-refusal.png?alt=media)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://help.metaforms.ai/tabulations/using-the-dialing-agent-for-quick-table-setup.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
