Connect your GitHub repos and let a local AI model transform your commits into clear, reader-friendly changelogs. Your source code never touches an external server.
We built ShipChanges with a privacy-first architecture. Here's the data flow:
Fetch
Diffs fetched from GitHub into memory
Filter
Sensitive files stripped automatically
Process
Local AI generates the summary
Save
Only the markdown summary is stored
No source code, diffs, or patches are ever persisted. Raw data lives only in RAM during processing.
Everything you need to ship professional changelogs, without compromising on security.
Your code never leaves your infrastructure. We use a local AI model running on your own hardware — zero data sent to external APIs.
We only save the generated changelog summary. Raw diffs, patches, and source code are processed in memory and immediately discarded.
Changelog generation runs through a local AI model on your server. Your code stays in RAM, never touches a third-party service.
Goes beyond commit messages. Reads actual code diffs to understand what changed, filtering out sensitive files like .env and credentials.
Set a daily, weekly, biweekly, or monthly schedule. Changelogs are generated automatically and queued for your review before publishing.
Brand your public changelog with your logo, colors, and display name. Each project gets its own shareable changelog page.
Connect your repos, configure your schedule, and let AI do the writing — all without your code ever leaving your machine.