Skip to content

Security: holashawty/ai-engineering-control-plane

Security

SECURITY.md

Security Policy

Reporting a vulnerability

If you discover a security issue in this repository (the framework itself, its generated entrypoints, or its evidence/memory tooling), please open a private security advisory or contact the maintainers directly rather than filing a public issue.

Handling secrets

  • Never commit credentials (tokens, API keys, passwords, private keys) to this repository, in code, configuration, or documentation.
  • If a secret is accidentally exposed anywhere (commit history, issue, chat log, CI log), rotate/revoke it immediately at the source (e.g. the provider's token management page) and remove it from history.
  • .aiecp/policy.local.yaml (if used) is git-ignored by default because it may contain locally pre-authorized, potentially sensitive automation scopes.

Scope of this document

See docs/security-model.md for the framework's full threat model, including credential handling, destructive-operation safety gates, AI output validation, supply-chain considerations, and memory-poisoning mitigations.

There aren't any published security advisories