> For the complete documentation index, see [llms.txt](https://docs.cherryai.com.cn/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.cherryai.com.cn/docs/en-us/advanced-basic/agent-workspace/workspaces-tasks-files.md).

# Working Directory, Tasks, and Files

An Agent’s “memory” and “files” are not the same thing. Tasks store the conversation process, the working directory stores the files that can be directly handled in this task, and Agent memory retains stable facts and experience across different tasks.

## Which working directory to choose

When starting a new task, the working directory selector provides existing directories, \[Add new working directory], and \[Do not use a working directory].

| Task type                                           | It is recommended to choose                  | Reason                                                                   |
| --------------------------------------------------- | -------------------------------------------- | ------------------------------------------------------------------------ |
| Modify code, organize materials, generate documents | A directory containing only this project     | Makes it easier for the Agent to find files and control the access scope |
| Temporary Q\&A or no local files needed             | \[Do not use a working directory]            | Reduce unrelated files from entering the task                            |
| Need temporary outputs but no existing directory    | Use the default workspace created by the app | Keep outputs separate from personal directories                          |

{% hint style="danger" %}
Do not choose the Desktop root directory, the user home directory, or a directory containing a large amount of private data just for convenience. The working directory determines the range of files the Agent can operate on, and the permission mode determines whether operations require confirmation; check both together.
{% endhint %}

<figure><img src="/files/aecf93d13d04d1712557ff1141528696395346ee" alt="Agent 工作区中的任务列表、工作目录和主操作区"><figcaption><p>Before starting, confirm the current Agent, task, and working directory to avoid writing files to the wrong project.</p></figcaption></figure>

## Manage tasks

* Click \[New task] to start a new context;
* Tasks can be renamed, pinned, and grouped by working directory;
* Right-click a task to choose \[Open in new tab] or \[Open in new window];
* An Agent keeps at least one task;
* When deleting a working directory record, the task records under it will be deleted, but the disk directory itself will not be deleted.

## Handle files in the right panel

After the Agent generates or modifies files, open \[Files] on the right to browse the workspace. Editable UTF-8 text files support switching between preview and edit mode; changes are automatically saved after input, so there is no need to click a separate save button.

If a file on disk is modified by another program during editing, Cherry Studio will prompt you to reload it. Choosing reload will discard the current draft; when keeping the draft, auto-save will pause until the conflict is resolved.

## Outputs and task status

The right-side \[Status] panel summarizes the outputs, sub-agents, workflows, background commands, and context usage declared by the Agent. When a long task is running, you can use this to first determine whether it is still working, waiting for permission, or has already failed.

<figure><img src="/files/cde6077f074e8ee59ce67e0bd26d30a8aa55bfa5" alt="Agent 右侧状态面板中的产物、任务与后台执行信息"><figcaption><p>For long tasks, first check [Status] to see whether it is running, waiting for approval, or has already failed.</p></figcaption></figure>

## What counts as completion

* The top of the task shows the directory for this project;
* You can find the agreed delivery files in \[Files] on the right;
* There are no pending approval or failed steps in \[Status] on the right;
* The original materials have not been overwritten, and the delivery files can be opened independently.

## User case: organizing interview materials

Create a new task for the “interview organization” Agent, and select only the folder for this interview as the working directory. Ask the Agent to keep the original records and create new `summary.md` and `quotes.md`. After the task is complete, check the citations and text in \[Files] on the right, and directly revise a few typos; for the next interviewee, open a new task to avoid mixing contexts.

<details>

<summary>Why is there no [Files] on the right?</summary>

The current task has no accessible working directory, or the directory is no longer accessible. First reselect the working directory at the top of the task, then send a message for the Agent to read the files.

</details>

<details>

<summary>Which files cannot be edited directly?</summary>

Binary files, non-UTF-8 text, and large files are only available for preview or external opening. They can still be processed by Agent tools, but they may not be editable in the built-in editor.

</details>


---

# 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, and the optional `goal` query parameter:

```
GET https://docs.cherryai.com.cn/docs/en-us/advanced-basic/agent-workspace/workspaces-tasks-files.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.
