Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

Sign up
Appearance settings

Security: agent-policy-protocol/spec

Security

SECURITY.md

Security Policy

Reporting a Vulnerability

If you discover a security vulnerability in the Agent Policy Protocol specification, reference implementations, or SDK, please report it responsibly.

Do NOT open a public GitHub issue for security vulnerabilities.

How to Report

Send an email to security@agentpolicy.org with:

  • A description of the vulnerability
  • Steps to reproduce the issue
  • The potential impact
  • Any suggested fixes (optional)

What to Expect

  • Acknowledgment: We will acknowledge receipt within 48 hours.
  • Assessment: We will assess the severity and impact within 5 business days.
  • Fix: We will work on a fix and coordinate disclosure with you.
  • Disclosure: We will publicly disclose the vulnerability after a fix is available, crediting you (unless you prefer anonymity).

Scope

This security policy covers:

  • The APoP specification (spec/)
  • The JSON Schema (spec/schema/agent-policy.schema.json)
  • The Node.js SDK (sdk/node/)
  • The reference middleware (middleware/)
  • The conformance test suite (tests/)

Supported Versions

Version Supported
1.0.x
< 1.0

Security Best Practices for Implementors

When implementing APoP in production:

  1. Always validate agent-policy.json against the JSON Schema before loading.
  2. Use HTTPS for all policy URLs, verification endpoints, and registry lookups.
  3. Implement rate limiting with a proper backend (in-memory counters, Redis, etc.) — the SDK's rate limit headers are advisory only.
  4. Verify agent signatures cryptographically — the SDK currently checks for header presence only, not cryptographic validity.
  5. Sanitize inputs — agent headers (Agent-Name, Agent-Intent, etc.) are untrusted input.

There aren’t any published security advisories

AltStyle によって変換されたページ (->オリジナル) /