Please report security vulnerabilities privately via GitHub's Security Advisory feature:
- Go to the Security tab
- Click "Report a vulnerability"
- Describe the issue, its impact, and any proof-of-concept
We'll respond within 48 hours. Do not open public issues for security vulnerabilities.
Only the latest version on main receives security updates.
Market Swarm sends your product YAML to the LLM provider you configure (Gemini, Anthropic, OpenRouter, or OpenAI). Never include trade secrets, real company data, or identifiable information in your product definitions.
The product YAML is the only data sent to cloud LLM APIs. All persona text, evaluation logic and results processing stay local.
If you're running simulations with sensitive product data:
- Use a private LLM provider (self-hosted Ollama, on-prem Claude models, private OpenRouter keys)
- Anonymize product names and competitor references
- Verify your LLM provider's data retention and compliance policies
This project covers simulation and reporting logic. Security issues outside scope:
- Vulnerabilities in upstream LLM providers (report directly to them)
- Vulnerabilities in your LLM key management (use environment variables, never hardcode)
- Issues with third-party package dependencies (report to the maintainers; we'll update pinned versions)
- EU AI Act Art. 50: Every output carries an AI-generation disclaimer
- GDPR: Product YAML data handling is transparent; no hidden data flows
- Python version: We support Python 3.11+ with regular dependency updates
- Support end-of-life Python versions
- Guarantee backward compatibility across major versions
- Review your LLM provider's terms; that's your responsibility