Skip to content

Security: mccarthy606/claude-operator-stack

Security

SECURITY.md

Security Policy

Reporting a vulnerability

If you find a vulnerability (a leaked secret in a config example, a curl | bash pattern that could be abused, an install script that does something unexpected, or any other security concern), please open a private security advisory through GitHub's "Security" tab on this repo.

If GitHub Security Advisories are not available to you, email the maintainer (handle in the LICENSE / README). Use a subject prefix of [security].

Please do not open a public issue for security concerns. The repo is small enough that a private heads-up is faster and reduces exposure.

What is in scope

  • Real secrets accidentally committed (API keys, tokens, OAuth credentials, private keys)
  • Install script behavior that runs with elevated privileges or modifies files outside expected paths
  • Hook configurations that could be exploited as a code-execution vector
  • Outdated or insecure dependency pins in any example config
  • MCP server examples that point to untrusted hosts

What is out of scope

  • General security advice for components this stack depends on (file those upstream)
  • Hardening recommendations for someone's own private setup (the stack is opinionated, not prescriptive)
  • Theoretical attacks against the broader Claude Code ecosystem — these belong upstream

Response time

Maintainer is solo. Expect an acknowledgement within 72 hours and a fix or mitigation plan within a week for confirmed issues. If you need urgency beyond that (active exploitation in the wild, real harm), say so explicitly in the report.

Disclosure

Once a fix is merged, the original reporter is credited in the changelog and (with permission) in credits/README.md. The vulnerability summary will be published as a GitHub Security Advisory.

There aren't any published security advisories