Security Policy
MINEAI-DEFENCE-SYSTEM (MADS) PROTOCOL
01. Security Architecture
MineAI utilizes a multi-layered security infrastructure known as MADS (MineAI-Defence-System). Our architecture is designed to protect user data, prevent API abuse, and ensure the integrity of our reasoning engines.
Encrypted Transit
All data is encrypted via TLS 1.3 in transit and AES-256 at rest via our Convex and Clerk infrastructure.
Node Isolation
High-risk tasks (Code Execution, Terminal Access) are run in isolated, stateless sandboxes with real-time revocation capabilities.
02. Responsible Disclosure
We welcome security researchers to help us strengthen our system. If you believe you've found a vulnerability, please follow our disclosure protocol:
- Do not attempt to access or modify data belonging to other users.
- Do not perform DoS or DDoS attacks on MineAI infrastructure.
- Provide detailed steps to reproduce the vulnerability.
- Allow us reasonable time to resolve the issue before public disclosure.
03. MADS Enforcement
The MADS v6.0 system monitors prompts and heuristics in real-time. Violations trigger automatic escalations:
Violation Penalties
Attempts to jailbreak models, bypass rate limits, or exploit the Deep Search protocol will result in immediate Risk Level escalation, ranging from temporary feature lockdowns to permanent account revocation.