Skip to content

Security: SwarmProtocol-fun/agent-guild

Security

SECURITY.md

Security Policy

Supported Versions

Version Supported
Latest Yes
< Latest Case-by-case

Reporting a Vulnerability

DO NOT open a public GitHub issue for security vulnerabilities.

Responsible Disclosure Process

  1. Email: ecosystemnetworkbayarea@gmail.com
    • Subject line: [SECURITY] Brief description
  2. Include:
    • Description of the vulnerability
    • Steps to reproduce
    • Potential impact assessment
    • Suggested fix (if any)
  3. You will receive an acknowledgment within 48 hours
  4. The team will provide a timeline for resolution within 7 days
  5. You will be credited (unless you request anonymity) once the fix is released

What to Expect

  • We take all reports seriously and will investigate promptly
  • We will not pursue legal action against researchers who follow this process
  • We will credit researchers in the release notes and CONTRIBUTORS.md

Bounty Program

The Agent Guild project allocates 1% of all project proceeds to the Security Bounty Reserve, as defined in the LICENSE and GOVERNANCE.md.

Bounty Tiers

Severity Description Bounty Range
Critical Remote code execution, key/credential exposure, full system compromise $5,000 — $50,000
High Authentication bypass, privilege escalation, significant data exposure $2,000 — $10,000
Medium Cross-site scripting, injection attacks, logic flaws with limited impact $500 — $2,000
Low Information disclosure, denial of service, minor misconfigurations $100 — $500
Informational Best practice violations, hardening suggestions Recognition + swag

Bounty amounts scale with the Security Bounty Reserve balance. The ranges above are targets — actual payouts are determined by the Core Team (Tier 2+) based on severity, impact, and quality of the report.

Eligibility

  • You must follow the responsible disclosure process above
  • The vulnerability must be in Agent Guild project code (not third-party dependencies)
  • The vulnerability must not already be known or reported
  • You must not exploit the vulnerability beyond what is necessary to demonstrate it
  • You must not violate the privacy of users or disrupt production systems
  • Participants in the DAO at Tier 3 (Guardian) and above are eligible but payouts may be adjusted to avoid conflicts of interest

Scope

In scope:

  • All code in the Agent Guild repository (excluding vendored/node_modules)
  • Smart contracts and blockchain integrations
  • API endpoints and authentication systems
  • Data storage and encryption implementations
  • Infrastructure-as-code configurations

Out of scope:

  • Third-party dependencies (report upstream, notify us)
  • Social engineering attacks
  • Physical security
  • Denial of service via brute force / volumetric attacks
  • Issues in development/staging environments only

Security Practices

For Contributors

  • Never commit secrets, API keys, or credentials
  • Use environment variables for sensitive configuration
  • Follow the principle of least privilege in all code
  • Validate all inputs at system boundaries
  • Keep dependencies updated and audit regularly

For the Core Team

  • Review all PRs for security implications
  • Run automated security scanning on all merges
  • Conduct quarterly security reviews
  • Maintain an incident response plan
  • Publish security advisories for resolved vulnerabilities

Past Advisories

No security advisories have been issued to date.


Contact

Security team: ecosystemnetworkbayarea@gmail.com

For non-security issues, use GitHub Issues.


This policy is part of the Agent Guild project and is subject to the terms of the Agent Guild Open Source License v1.0.

There aren't any published security advisories