| Version | Supported |
|---|---|
| 0.1.x | ✅ |
If you discover a security vulnerability in AlphaEvo, please report it responsibly.
Do NOT open a public issue.
Instead, report privately via GitHub Security Advisories with:
- A description of the vulnerability
- Steps to reproduce (if applicable)
- Impact assessment
We will acknowledge receipt within 48 hours and aim to provide a fix or mitigation within 7 days for critical issues.
AlphaEvo is a research tool that processes publicly available market data. It does not handle:
- Real trading or brokerage credentials
- Personal financial data
- User authentication tokens (beyond optional LLM API keys)
Security concerns most relevant to this project:
- LLM prompt injection — Malicious strategy descriptions could attempt to manipulate LLM behavior. The parser validates all LLM output against the Strategy DSL schema before execution.
- Arbitrary code execution — Alpha Factory sandboxes synthesized factor code. Only whitelisted operations are permitted.
- Dependency supply chain — We pin transitive dependencies and run
pip-auditin CI (planned). - SQL injection — All database access uses parameterized queries via
Python's
sqlite3module.
We follow coordinated disclosure. Credit will be given to reporters unless anonymity is requested.