GitHub Desktop
Built on Electron, GitHub Desktop stores Chromium disk cache entries in Cache/Cache_Data for web content used by its interface and authentication flows, and on Windows it also keeps a GPUCache with compiled rendering data such as shader and GPU resource cache files. These files speed up repeat loads of embedded web views, avatars, sign-in pages, and hardware-accelerated UI rendering, but they can become stale after app or graphics driver changes. Kudu removes only those disposable cache files so GitHub Desktop can rebuild them, without touching repositories, credentials, account data, or preferences.
Why clean GitHub Desktop?
- Corrupted Chromium disk cache entries can make sign-in and OAuth web views load incorrectly, leaving you with blank pages, missing buttons, or login loops
- Stale cached avatars, issue links, and embedded web content can cause parts of the interface to show outdated or broken images until the cache is rebuilt
- After a GitHub Desktop update, invalid cached web assets can trigger slow first launches, incomplete rendering, or panels that appear empty before refreshing
- On Windows, GPUCache data from an older graphics driver can be invalidated after a driver update, causing flicker, black rectangles, or sluggish rendering in hardware-accelerated views
- An oversized Cache_Data folder can waste hundreds of megabytes on repeated network resources, especially if the app has been used heavily across many repositories and account sessions
- Damaged cache files can lead to inconsistent interface behavior such as delayed repository list updates, missing commit history visuals, or repeated reloads of embedded pages
Cache paths Kudu targets
Windows
%AppData%/GitHub Desktop/Cache/Cache_Data |
%AppData%/GitHub Desktop/GPUCache |
macOS
~/Library/Application Support/GitHub Desktop/Cache/Cache_Data |
Common questions about GitHub Desktop
Download Kudu and reclaim your disk space.
Available on Windows, macOS, and Linux. No account required, no feature gates, no telemetry without consent. All cleaning targets are open source and community-auditable.