Security
Last updated: November 18, 2026
Steward holds your plans, budgets and team details, so we build with security in from the start. This page describes the measures in place today. We're a small team and don't hold certifications such as SOC 2 yet; we rely on providers that do, and we'll keep this page current.
Access control in the database
- Every table uses Postgres row-level security. Each read and write is checked by the database against your project membership, so a bug in the app can't show one customer's projects to another.
- Only a project's owner can manage its members, integrations and deletion. Members can't change their own role or take over a project.
- Plan and billing fields can only be changed by our payment webhook, never from the browser, so no one can grant themselves a paid plan.
- Slack and Discord webhook URLs are secrets: only the project owner can see them.
Infrastructure and encryption
- The app is hosted on Vercel and the database and authentication run on Supabase, in the United States. Both providers hold SOC 2 Type 2 reports.
- All traffic is encrypted with HTTPS (TLS). Data is encrypted at rest by our database provider, including backups.
- Database keys with elevated access are used only on our servers, never sent to the browser.
Sign-in
- Passwords are stored only as salted one-way hashes by our authentication provider. We can't see them.
- You can sign in with Google, GitHub or an emailed one-time link instead of a password.
- Changing your password requires your current password.
- Sign-in redirects only go to pages on our own site, so links can't be used to send you elsewhere.
Payments
Payments are handled by Stripe, a PCI DSS Level 1 certified provider. Card details are entered on Stripe's own pages and go straight to Stripe: they never touch our servers, and we never see your full card number. Payment events from Stripe are verified with a signature before we act on them.
Application protections
- Requests that change data from another website are rejected, protecting you against cross-site request forgery.
- Integrations only send to official Slack and Discord addresses, so our servers can't be pointed at other systems.
- CSV exports are protected against spreadsheet formula injection.
- Our pages send security headers that restrict framing and limit what scripts can run.
Your data, your control
- Export your data from Settings at any time.
- Delete your account in Settings. This deletes every project you own, removes you from other projects and cancels any subscription. Backups are overwritten within 30 days.
- We only use strictly necessary cookies. See the Cookie Policy.
Incidents
If we learn of a security incident that affects your personal data, we'll notify affected users without undue delay, and the relevant authorities where the law requires (for example, within 72 hours under GDPR).
Reporting a vulnerability
If you think you've found a security issue, email security@example.com with steps to reproduce it. We'll acknowledge your report within 3 business days and keep you updated while we fix it.
We won't take legal action against good-faith research that:
- only uses accounts you own or have permission to test with;
- doesn't access, change or delete other people's data;
- doesn't degrade the service for others (no denial-of-service or spam testing);
- gives us reasonable time to fix the issue before you share it publicly.
We don't run a paid bug bounty at the moment, but we're happy to credit you publicly if you'd like.