Security at SkillRadar
Your skill data is sensitive. We treat it that way.
Infrastructure
- All data encrypted in transit (TLS 1.3) and at rest (AES-256)
- Hosted on Vercel (frontend) and managed PostgreSQL with automated backups
- No customer data stored in browser local storage or cookies beyond session tokens
- Regular dependency audits and security updates
Authentication
- Passwordless authentication: no password storage, no credential stuffing risk
- Social OAuth (Google, GitHub, LinkedIn, Microsoft) with magic link fallback
- Database-backed sessions with server-side validation and revocation
- Enterprise SSO (SAML/OIDC) available on Enterprise plan
Data privacy
- Peer ratings are never exposed individually. Aggregates only, with minimum response thresholds
- All API inputs validated with strict schema enforcement
- AI suggestion data retained for 30 days only, then permanently deleted
- Full GDPR compliance: export and delete your data at any time
AI safety
- All AI-generated content goes through a human review queue and is never written directly to live data
- External AI integrations (MCP, API) write to a suggestion queue, not production data
- AI providers (Anthropic) do not use your data for model training
Reporting vulnerabilities
If you discover a security vulnerability, please report it responsibly to security@skillradar.com. We take all reports seriously and will respond within 48 hours.