Skip to main content
Trust Center

Security practices

The concrete controls that protect your account and your data. Each is mapped to our internal security system, which labels anything it cannot verify rather than assuming it is safe.

Identity & access

  • Least-privilege, allowlisted admin access

    In place

    Internal tools are restricted to an explicit operator allowlist with role-based permissions, so a single account cannot change everything.

Data protection

  • Secrets never reach the browser

    In place

    API keys and service credentials live only in server environments and are never included in pages delivered to your browser.

  • Database row-level isolation

    In place

    When cloud accounts are enabled, row-level security ensures you can only read and write your own data — enforced at the database, not just in the app.

  • On-device video analysis by default

    In place

    Swing videos are analyzed in your browser. Footage is not transmitted to external servers by default, and is never used to train AI models without your explicit consent.

Application security

  • HTTPS everywhere with hardened security headers

    In place

    Every page is served over HTTPS with a Content-Security-Policy, X-Frame-Options, and HTTP Strict-Transport-Security to defend against injection, clickjacking, and protocol downgrade.

  • Validated, bounded uploads

    In place

    Uploads are restricted to known media types with enforced size and duration limits, reducing the risk of malformed or abusive files.

  • User-generated content is sanitized

    In place

    Anything you type that is shown back in the interface is sanitized before it is rendered.

AI safety

  • AI abuse & spend safeguards

    In place

    AI endpoints sit behind rate limiting and a global daily spend cap, so abuse cannot degrade service or produce a runaway bill.

Monitoring & incident response

  • Admin audit logging

    In place

    Sensitive internal actions are recorded so changes are accountable and incidents can be reconstructed.

  • Written incident-response runbook

    In place

    We maintain a runbook covering secret exposure, account compromise, and data-leak scenarios so response is a checklist, not a scramble.

Secure development

  • Automated security scanning in CI

    On roadmap

    We are rolling out secret scanning, dependency vulnerability scanning, and static analysis on every code change. Remaining gaps are tracked openly.

Found a security issue? Please report it responsibly via our vulnerability disclosure policy.