---
title: "Projects"
url: "https://docs.mekodata.ai/use-meko/projects/"
---

# Projects

Group datapacks that work toward one outcome

A project groups datapacks that work toward one outcome. Use projects to give related datapacks a visible structure without combining them into a single store. Each datapack remains an isolated home for its conversations, memory, and knowledge.

A project is structured using folders that you can use to group datapacks that do similar work together. The structure has three levels:

- The project represents the outcome.
- Each folder represents one kind or stage of work.
- Each datapack in a folder holds the focused context for that work.

Folders do not nest. A datapack can belong to only one project. Datapacks that are not in a project remain **Unfiled** and continue to work normally.

## Sharing project datapacks

Each datapack in a project is shared on its own, so two people can see different parts of the same project, and the same person can have a different role on each datapack you share with them.

When you share a datapack that has been added to a project, the recipient finds its project on their **Projects** page under **Shared with me**. They can see the datapack and the folder it sits in. The other folders and datapacks in the same project stay hidden. **Add folder**, **Add a datapack**, and **Settings** stay with the project owner.

## Create a project

To create a project:

1. On the **Projects** page, click **New Project**.
2. Enter a name and, optionally, a description.
3. Choose a layout. Built-in layouts already include their folders. **Custom** starts with no folder, so name the first one.
4. Click **Create project**.

### Add folders and datapacks

To add folders to a project:

1. Open the project and choose **Folders**.
2. Click **Add folder**.
3. Enter a name and, optionally, a description.
4. Click **Add folder**.

To add a datapack to a folder, select a folder in the **Folders** list and click **Add a datapack**.

You can also add a datapack to a folder from the datapack page:

1. Go to the datapack, click the actions menu (the gear next to **Share**), and choose **Add to project**. If the datapack is already in a project, the option is **Move to project**.
2. Select the project and folder.
3. Click **Add here** (or **Move here**).

### View project stats and activity

Under **Folders**, select **learnings** to view memories from the datapacks in the project, including ones still waiting for a promotion decision.

Below the folder list, **What this project holds** shows the project's totals.

Choose **Activity** to see a log of the files uploaded, memories promoted, folders created, and datapacks filed in the project.

## Manage a project from an agent

An agent on Meko Cloud can create, list, update, and delete projects, manage folders, add or remove datapacks, and read project stats.

Deleting a project or a folder from the agent keeps the datapacks, the same as in the portal. The layout is gone, so confirm before the agent deletes it.

Only the owner can change a project. A project shared with you answers as not found. Do not create another project with the same name. That makes a second project in your own tree and leaves the shared one as it is.

Filing a datapack is that datapack's owner's choice. A datapack shared with you cannot be filed into your project.

## Delete a project or a folder

If you own the project, open **Settings** to delete it. The project and its folder structure are removed for good. The datapacks filed inside are not deleted: they become **Unfiled** and keep working, with their conversations, memory and knowledge untouched. File them into another project whenever you like.

Deleting a folder works the same way, and leaves the rest of the project in place: the folder is removed, and the datapacks inside it become **Unfiled**. There is no button for this in the app yet — folder deletion is available over the API and through the MCP project tools.

To delete a datapack itself, open that datapack and delete it there.