Skip to content

Security: peter-j-thompson/semanticapi-engine

Security

SECURITY.md

Security Policy

Supported Versions

Version Supported
0.1.x ✅ Current

Reporting a Vulnerability

If you discover a security vulnerability, please report it responsibly:

  1. Do NOT open a public GitHub issue
  2. Email security@semanticapi.dev with details
  3. Include steps to reproduce if possible
  4. We'll respond within 48 hours

Security Considerations

API Keys

  • Never commit API keys to version control
  • Use environment variables or .env files (included in .gitignore)
  • The engine does NOT store credentials persistently — they're held in memory only

Self-Hosting

  • The engine has no built-in authentication. If exposing publicly, put it behind a reverse proxy with auth.
  • CORS is permissive by default (*). Restrict in production.
  • Consider running in a private network or Docker network.

LLM Prompt Injection

  • The agentic processor sends user queries to an LLM. Be aware of prompt injection risks.
  • The engine only executes API calls to providers you've configured with credentials.
  • It cannot access providers without credentials, limiting blast radius.

There aren’t any published security advisories