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

# Drives

> Organize files by project, agent, or team

A **drive** is a container for files. Think of it like a folder or a project — each drive isolates its files from other drives.

## When to create a drive

* One drive per project (`research-q1`, `marketing-assets`)
* One drive per agent (`data-analyst`, `code-reviewer`)
* One drive per team (`engineering`, `design`)

## Drive slugs

Claimed drives get a human-readable slug: `username/drive-name`. For example, `alice/research-q1`.

Slugs can be used as an alternative to UUIDs wherever `drive_id` is accepted:

```bash theme={null}
# Using UUID
GET /v0/files?drive_id=d7f3e-abc-123

# Using slug
GET /v0/files?drive_id=alice/research-q1
```

You can also look up a drive by slug:

```bash theme={null}
GET /v0/drives?slug=alice/research-q1
```

Unclaimed drives (provisioned by agents but not yet claimed) have no slug — only a UUID.

## Drive lifecycle

| State     | Duration      | Behavior                               |
| --------- | ------------- | -------------------------------------- |
| Unclaimed | Up to 30 days | Full read/write, 100 MB limit, no slug |
| Claimed   | Permanent     | Plan-based limits, slug active         |

Unclaimed drives auto-delete after 30 days. Claiming a drive (via the claim URL) removes the expiry and activates the slug.

## Drive limits

| Plan | Max drives |
| ---- | ---------- |
| Free | 3          |
| Pro  | Unlimited  |
| Team | Unlimited  |

## Drive metadata

Each drive tracks:

* `file_count` — Number of files in the drive
* `total_size` — Total storage used (bytes)
* `slug` — Human-readable identifier (null if unclaimed)
* `expires_at` — Auto-deletion date (null if claimed)
* `created_at` / `updated_at` — Timestamps

These counters update automatically as files are uploaded or deleted.