The Xclaude Gateway
Manifesto

We didn't come to this as engineers. We came as two people who love what Claude can do — and who, while learning to build with it, kept noticing how easily a small mistake could become a real one.

The moment you connect Gmail, your calendar, your files, an AI agent stops being a chat window and starts being a set of hands. Most of the time, that's a gift. But hands you can't watch are hands you can't trust. And today, most people plugging tools into AI have no way to see what's actually happening underneath.

That gap is why Xclaude Gateway exists.

We believe transparency is the foundation of trust. You should be able to see every action an AI takes on your behalf — not a summary, not a promise, the actual calls. Xclaude sits on your computer and routes the connections between Claude and your tools through a window you control. Nothing is hidden.

We believe your data is yours. Xclaude runs entirely on your machine. Nothing leaves it — no logs, no telemetry, no exceptions. We practice what we preach: if we ask you to trust an open, local tool, we won't quietly ship your data somewhere else.

We believe safety should come before the mistake, not after. New users shouldn't have to learn the hard way. By making every action visible and logged, Xclaude helps you catch a critical error — or a malicious one — before it costs you something.

We believe the best safeguards are open. Xclaude is open source because security you can't inspect isn't security. Read it, audit it, improve it, fork it.

We built this for one simple reason: to make Claude safer to use, for everyone still learning how.