> For the complete documentation index, see [llms.txt](https://docs.coherent.global/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.coherent.global/assistant/agent/what-is-agent.md).

# What is Agent?

Agent is an AI assistant built into Coherent Assistant for Excel. It helps you understand workbooks, explain formulas, analyze data, create pivot tables, compare files, and complete everyday Excel tasks without leaving Excel.

<figure><img src="/files/oYPuhy1IutUhiJp0o5mp" alt="Agent working beside a spreadsheet workbook"><figcaption><p>Agent works beside your workbook to help explain, review, compare, and improve spreadsheet content.</p></figcaption></figure>

Agent works beside your workbook. It can use the active sheet, selected cells, attached files, saved outputs, and the conversation so far to answer questions or complete a task. When a request needs more information, Agent asks a follow-up question. When a request would change workbook content, Agent can ask for approval before it acts.

## What Agent is good for

Use Agent when you want to move from a workbook question to a practical next step. For example, Agent can help you:

* Understand how a workbook is organized and what each sheet is for.
* Explain formulas, assumptions, outputs, and calculation flow.
* Summarize sheets, ranges, tables, and attached data files.
* Find hardcoded values, inconsistent formulas, errors, external links, and other review signals.
* Update values, formulas, formatting, sheets, and ranges after you approve the change.
* Create, refresh, and adjust pivot tables.
* Compare workbook versions and summarize material differences.
* Use saved instructions for repeated team workflows.

Agent is most useful when you give it a clear area to inspect and a clear outcome. Instead of asking it to "analyze everything," start with a workbook, sheet, range, table, or attachment.

## How Agent works with your workbook

You chat with Agent in plain language. Agent reads the relevant workbook context from your device, decides what information it needs, and uses AI to produce a response. It does not upload your whole workbook just because you open Agent.

Agent can work in different levels of control. In read-only mode, it reviews and explains. In editing modes, it can prepare workbook changes and ask for approval before applying them. This helps you keep review work separate from workbook edits.

## Example requests

Try requests like these:

* "Summarize the active sheet and identify the main inputs and outputs."
* "Find formulas that reference the Assumptions tab."
* "Review this workbook for potential model issues and group findings by severity."
* "Create a pivot table from the selected range with Region as rows and Revenue as values."
* "Compare the attached workbook with the open workbook and list material differences."

If Agent's answer is too broad, ask a narrower follow-up. If the task is important, ask Agent to include cell references and assumptions.

## Safety and control

Agent is designed to keep you in control. Use read-only mode for review, Ask for edit mode when you want to approve changes, and Agent mode only for trusted tasks with clear instructions.

Review important AI outputs before relying on them, especially when they affect formulas, business assumptions, customer communication, or reporting. See [AI Guidelines](/assistant/agent/ai-guidelines.md) for privacy, data handling, and AI limitations.

## Related pages

* [AI Guidelines](/assistant/agent/ai-guidelines.md)
* [Getting started with Agent](/assistant/agent/getting-started-with-agent.md)
* [Chatting with Agent](/assistant/agent/chatting-with-agent.md)
* [Agent modes and approvals](/assistant/agent/agent-modes-and-approvals.md)
* [Using files and attachments](/assistant/agent/using-files-and-attachments.md)


---

# 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 dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.coherent.global/assistant/agent/what-is-agent.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.