If you discover a security vulnerability in FlowOS, please report it responsibly.
Do NOT open a public GitHub issue for security vulnerabilities.
Instead, please email: security@flowos.dev (or create a private advisory on GitHub)
Include:
- Description of the vulnerability
- Steps to reproduce
- Potential impact
- Suggested fix (if any)
- Acknowledgment: within 48 hours
- Initial assessment: within 1 week
- Fix or mitigation: depends on severity, but we aim for quick resolution
FlowOS is designed with security in mind:
- All data lives in a local SQLite database
- No data leaves your machine unless you configure a cloud AI provider
- API keys are stored in the database, not in environment variables or config files
- Every agent action is logged to an audit trail
- Risky operations require user confirmation (confirm tier)
- Destructive operations are blocked by default (blocked tier)
- Every action is undoable via the audit log
- The agent is scoped to visible context only — it cannot see data you're not viewing
- API keys are sent only to the configured AI provider
- No data is sent to third parties beyond the configured provider
- Provider connections can be validated before use
- Tool risk tiers can be overridden per-tool
- Users can raise friction (auto → confirm) freely
- Lowering friction requires staying within safe bounds
- Blocked tools can never go below confirm tier
This security policy applies to the FlowOS application code in this repository. It does not apply to:
- Third-party AI providers (Anthropic, Nvidia, OpenAI, etc.)
- Your operating system or network security
- Dependencies (though we monitor for known vulnerabilities)
We use npm audit to monitor for known vulnerabilities in dependencies. If you discover a vulnerable dependency, please report it.