Skip to content

Security: hcompai/agp-playground

SECURITY.md

Security Policy

Our Commitment

The security of AgP Playground is important to us. We appreciate your efforts to responsibly disclose any security vulnerabilities you find.

Reporting a Vulnerability

Please do not report security vulnerabilities through public GitHub issues.

How to Report

If you discover a security vulnerability, please send an email to:

privacy@hcompany.ai

Include the following information:

  • Type of vulnerability: (e.g., XSS, CSRF, credential exposure, etc.)
  • Full description: Detailed explanation of the vulnerability
  • Steps to reproduce: Clear steps to reproduce the issue
  • Proof of concept: Code sample or screenshots demonstrating the vulnerability
  • Impact: Potential impact if the vulnerability is exploited
  • Suggested fix: If you have a recommendation for fixing it
  • Your contact information: For follow-up questions

What to Expect

  1. Acknowledgment: We'll acknowledge receipt of your report within 48 hours
  2. Initial assessment: We'll provide an initial assessment within 5 business days
  3. Updates: We'll keep you informed of our progress
  4. Resolution: We aim to resolve critical vulnerabilities promptly
  5. Disclosure: We'll work with you on a coordinated disclosure timeline

Security Best Practices

For Users

1. API Key Management

Never commit API keys to version control:

Bad:

// .env.local committed to git
AGP_API_KEY=sk_prod_abc123...

Good:

// .env.local in .gitignore
AGP_API_KEY=sk_prod_abc123...

Best practices:

  • Store API keys in .env.local (not tracked by git)
  • Never share API keys in screenshots or issues
  • Use different keys for development and production
  • Revoke unused or compromised keys immediately at Portal-H

2. npm Access Token

Keep your npm token secure:

  • Store in environment variable (NPM_TOKEN)
  • Add .npmrc to .gitignore (already done)
  • Don't share tokens publicly
  • Revoke old tokens at npmjs.com/settings/tokens

3. Running Untrusted Agent Tasks

Be cautious when testing agents:

  • Review agent objectives before running
  • Understand what URLs the agent will visit
  • Be aware of sensitive data in your environment
  • Use test accounts when possible

For Contributors

1. No Secrets in Code

  • Never commit API keys, tokens, or credentials
  • Use .env.example files with placeholder values
  • Review commits before pushing
  • Check git history if you accidentally commit secrets

2. Secure Coding Practices

  • Sanitize user inputs
  • Validate environment variables
  • Use HTTPS for all API calls
  • Implement proper error handling (don't leak sensitive info)

3. Dependencies

  • Keep dependencies updated
  • Review security advisories for dependencies
  • Use npm audit or pnpm audit regularly

Common Security Scenarios

Scenario 1: Exposed API Key

If you accidentally expose an API key:

  1. Revoke immediately at Portal-H
  2. Generate a new key
  3. Update your .env.local
  4. Check git history - if committed, rotate the key
  5. Monitor for unauthorized usage

Scenario 2: Exposed npm Token

If you expose your npm token:

  1. Revoke immediately at npmjs.com/settings/tokens
  2. Generate a new token
  3. Update your environment: export NPM_TOKEN=new_token
  4. Remove from git history if committed

Scenario 3: XSS or Injection

If you find a potential XSS or injection vulnerability:

  1. Don't exploit it publicly
  2. Report to privacy@hcompany.ai
  3. Include steps to reproduce
  4. We'll work on a fix

Security Features

Built-in Protections

  • Environment variable validation: Required env vars are checked on startup
  • HTTPS enforcement: All API calls use HTTPS
  • No credential storage: API keys stored in browser localStorage (not accessible via XSS if properly configured)
  • Screenshot proxy: Authenticated screenshots use server-side proxy

What's NOT Protected

This playground is a development tool and should not be used in production as-is:

  • No rate limiting
  • No authentication system
  • API keys stored in browser
  • No audit logging

Scope

This security policy applies to:

  • The AgP Playground application
  • Example workflows
  • Documentation

Out of Scope

  • Vulnerabilities in third-party dependencies (report to dependency maintainers)
  • Vulnerabilities in the AgP SDK itself (report to SDK repository)
  • Vulnerabilities in the AgP platform (report separately)

Updates

Security updates will be announced through:

  1. GitHub Security Advisories
  2. GitHub Releases
  3. CHANGELOG.md

Questions?

If you have questions about this security policy, please contact us at privacy@hcompany.ai.


Last Updated: November 12, 2025

Thank you for helping keep AgP Playground and our users safe!

There aren't any published security advisories