The Valarchi GitHub app has read-only access to your repository. It cannot write, push, or modify your code in any way.
We do not store your source code. When you connect a codebase, Valarchi extracts a semantic snapshot — product context,information architecture, UI patterns and design tokens. The raw code is not retained.
Everything extracted from your codebase is visible to you in your product's details page. There is no hidden data — what you see is everything we have.
Agents work from your extracted snapshot and the planning artifacts you create — PRDs, app structure, wireframes, and tech plans.
All agent inputs and outputs are visible in the activity panel for each agent run. There are no invisible inputs — if the agent used it, you can see it.
Anthropic retains API inputs and outputs for 30 days under their standard API terms, after which they are deleted.
Your content is not used to train Claude. User feedback sharing with Anthropic is disabled on our account.
For full details on how Anthropic handles API data, see their Privacy Policy →
If you bring your own Anthropic API key, Your key is encrypted before it's saved and never shown again in Valarchi.
Only organization admins can add or remove the API key.
All data in Valarchi is scoped to your organization. Products, milestones, planning artifacts, and usage history are only accessible to members of your workspace.
Have a security question? Contact us →