Skip to content

Security: hermes-labs-ai/claude-router

Security

SECURITY.md

Security Policy

Scope

claude-router is a local prompt-routing utility. Security issues in scope include:

  • code execution bugs in the package
  • unsafe file handling
  • dependency vulnerabilities introduced by this repo
  • misconfigurations that could cause the CLI or library to expose secrets unexpectedly

Out of scope:

  • benchmark disagreements
  • workload-specific routing disagreements
  • bugs in external Claude or Ollama services

Reporting

Open a private security advisory on GitHub if possible. If that is not available, open an issue only for non-sensitive reports.

Safe usage notes

  • do not commit private prompts, API keys, or benchmark data
  • treat routing decisions as workload-dependent and validate them before production use

There aren't any published security advisories