SECURITY OVERVIEW

1. Architecture

The Service uses multi-tenant PostgreSQL with row-level security (RLS) so each account accesses only its own watches, snapshots, signals, and briefs.

Privileged keys for agent runs and LLM APIs are held server-side only; web and mobile clients authenticate with user-scoped tokens.

2. Authentication and access

Sign-in uses magic-link email authentication via Supabase Auth. Sessions use secure HTTP-only cookies on web and platform secure storage on mobile.

Internal access follows least-privilege principles.

3. Encryption

Data in transit is protected with TLS. Data at rest is encrypted by our cloud infrastructure providers. Secrets are stored as environment variables, not in source control.

4. Application controls

API routes validate input. Budget caps and pause controls limit runaway AI spend. Dependency updates and vulnerability monitoring are recommended for all deployments.

5. Monitoring and incidents

We log application errors, authentication events, and usage for reliability and security. Report suspected incidents to security@vigilant.app.

We will notify affected customers of breaches affecting personal data as required by law and our DPA.

6. Responsible disclosure

We welcome responsible vulnerability reports at security@vigilant.app. Please include reproduction steps and avoid accessing other customers' data.

7. Compliance roadmap

Formal certifications (e.g., SOC 2) may be pursued as the product matures. Contact legal@vigilant.app for enterprise security questionnaires.