| Version | Supported |
|---|---|
| Latest | ✓ |
If you discover a security vulnerability in ODEI or its API, please report it via:
- Email: security@odei.ai
- GitHub Security Advisories: https://github.com/odei-ai/web/security/advisories
We take security seriously and will respond within 48 hours.
ODEI's constitutional world model includes built-in security measures:
- 7-layer constitutional validation before every agent action
- Provenance verification to prevent prompt injection
- Authority checking to prevent privilege escalation
- Rate limiting and JWT authentication on all API endpoints
- ERC-8004 on-chain agent identity for trust verification
- Security: security@odei.ai
- Product: api.odei.ai
- Twitter: @odei_ai