Skip to content
This repository was archived by the owner on Oct 22, 2026. It is now read-only.

Security: agentmiddleware/agent-middleware-api

Security

SECURITY.md

Security Policy

Supported Versions

Version Supported
main (latest) Yes
older branches/tags No

Reviewing This Repository

Attacking the trust plane deliberately? docs/security-review-kit.md is the rules of engagement: which target to use, how to provision your own credentials, the invariants a finding has to break, and the limits that are already documented in SECURITY_LIMITATIONS.md.

Reporting a Vulnerability

If you discover a security issue:

  1. Do not disclose it publicly in Issues or Discussions.
  2. Open a private security advisory in GitHub (preferred).
  3. If advisory flow is unavailable, open an issue with minimal details and request private contact.

Include:

  • affected endpoint(s)
  • reproduction steps
  • impact assessment
  • proposed mitigation (if available)

Response Expectations

  • Initial triage target: within 5 business days
  • Remediation timeline: based on severity and exploitability
  • Credit: provided upon request after a fix is published

There aren't any published security advisories