Skip to content

Security: yogayulanda/forge-context-engine

SECURITY.md

Security Policy

Forge Context Engine is a documentation and context-lifecycle foundation. It is not a deploy platform, runtime execution platform, CI/CD runner, or autonomous agent system.

Reporting Security Concerns

If you find a security issue, please report it privately to the repository maintainer.

Do not open a public issue that includes:

  • production credentials
  • API keys, tokens, private keys, cookies, or passwords
  • credential-bearing URLs
  • sensitive customer data
  • internal incident details that should not be public

Use sanitized examples whenever possible. If a secret may have been exposed, rotate it before sharing any reproduction details.

What To Report

Security and governance concerns are welcome, including:

  • accidental secret or PII exposure in examples, docs, context, or artifacts
  • unsafe guidance around financial correctness, idempotency, retry/replay, rollback, or validation honesty
  • wording that could imply autonomous approval of high-risk decisions
  • documentation that appears to introduce runtime execution, deploy automation, or hidden persistent memory behavior

Responsible Disclosure Expectations

  • Give maintainers reasonable time to investigate and patch.
  • Share only the minimum evidence needed to reproduce or understand the issue.
  • Avoid publishing exploit details or sensitive examples before maintainers respond.

Forge's security posture is intentionally lightweight and practical: prevent unsafe context handling, avoid sensitive data exposure, and preserve human approval for high-risk engineering decisions.

There aren't any published security advisories