If you discover a vulnerability, ethical concern, or misuse risk related to this project, please report it responsibly.
Email reports to:
eddy.projectvirgil@proton.me
We aim to respond to all security-related messages within 5 business days.
Please include as much relevant information as possible, including:
- A clear description of the issue
- Steps to reproduce (if applicable)
- Potential impact
- Any suggested mitigations
This project includes experimental tools designed to function in degraded, post-collapse environments. While we prioritize safety and ethical operation, SHEPHERD is provided as-is, without warranties or guarantees.
Known scope areas include:
- Offline use of local AI models
- Interaction with survival and ethical training datasets
- Limited execution environments (Raspberry Pi, air-gapped systems, etc.)
This project does not include:
- Online data fetching
- Remote telemetry
- Centralized model governance
SHEPHERD is designed to be independently auditable and self-contained.
Contributors and users are expected to uphold the core principles of the SHEPHERD system:
- Do no harm
- Preserve human dignity
- Encourage ethical survival
- Remain transparent and teachable
If any use of this system violates those principles, it is considered misuse.
Thank you for helping us keep SHEPHERD safe and aligned.