Privacy & Data Sharing

When you browse a website with unbrowse, the API structure of that site is automatically indexed into the collective registry. This includes:

• API endpoint URLs and HTTP methods (e.g. GET /v6/quote)
• Request and response schemas (parameter names, types, shapes)
• URL patterns and path templates
• Content types and header patterns (not values)
• General endpoint behavior (idempotency, pagination patterns)

Think of it like contributing a map. You're sharing the shape of the roads, not what's in anyone's car.

Your personal and authentication data is never shared, uploaded, or transmitted. This includes:

• Cookies, session tokens, and auth headers
• API keys, bearer tokens, and credentials
• Usernames, passwords, and account information
• Request/response bodies containing personal data
• Any data you send to or receive from APIs
• Your browsing history and session recordings

Credentials are stored in a local encrypted vault on your machine. When you replay a skill, your auth is injected locally — it never passes through our servers.

Unbrowse collects anonymous usage telemetry to improve the product. This includes:

• Which skills are replayed and how often (aggregate counts, not per-user)
• Replay success/failure rates per endpoint
• Skill discovery and indexing events (what domains, not who)
• Error types and crash reports (no personal context)
• Client version and platform (OS, runtime)

Telemetry is used to calculate reliability scores, route agents to the best endpoints, and prioritize improvements. It is never sold or shared with third parties.

To be clear: unbrowse does not collect, process, or store personal data as defined by GDPR, CCPA, or similar regulations. The data shared with the collective index is website infrastructure information — the same endpoint structures any developer could discover by opening browser DevTools.

We're indexing how websites work, not who uses them.

You can disable telemetry at any time:

You can also run unbrowse in local-only mode, which skips indexing to the shared registry entirely. Skills you discover stay on your machine.

If you have questions about what data unbrowse handles, reach out at @getFoundry or open an issue on GitHub.

Unreel AI Pte Ltd · Singapore