Security is a top priority for this project. Since this repository curates templates, instructions, and tools for autonomous AI agents, we take the integrity of our links and configuration files very seriously.

We apply security patches and updates primarily to the main branch of this repository. If you are using our templates or configuration files, please ensure you are pulling from the latest commit on main.

For this specific repository, we consider the following to be critical security vulnerabilities:

1. Malicious Links: Any curated link pointing to malware, phishing sites, or compromised third-party AI tools.
2. Harmful Prompts/Injections: System prompt templates or .cursorrules that contain hidden instructions to leak API keys, exfiltrate data, or execute unauthorized code.
3. Compromised Integrations: Listed MCP servers, AI plugins, or API endpoints that have been taken over by malicious actors.
4. Credential Leaks: Accidental commits containing active API keys, OpenAI tokens, or Anthropic keys.

Note: Bugs or typos in formatting that do not pose a security risk should be submitted as a standard Pull Request or Issue, not a security report.

Please do not report security vulnerabilities through public GitHub issues.

If you believe you have found a security vulnerability, please report it to us privately:

1. Email: Send a detailed report to [security@yourdomain.com].
2. GitHub Private Reporting: Alternatively, use the "Report a vulnerability" button under the Security tab of this repository.

• The exact file, link, or template containing the vulnerability.
• A description of the potential exploit (e.g., "The XYZ MCP server link redirects to a known malware domain").
• (Optional) Suggested remediation.

We will acknowledge receipt of your vulnerability report within [48 hours] and strive to resolve or patch the issue within [7 days]. Once the issue is resolved, we will publish a post-mortem or advisory if the vulnerability impacted users.

Thank you for helping keep the agent ecosystem safe and secure.