✨ Feature Request: Add automated safety scanning for new AI skills via CI/CD

[ashp15205]

Hey team! This repository is an incredible resource. Since you are aggregating hundreds of cybersecurity skills for AI agents, there is an inherent risk of community submissions accidentally (or intentionally) containing cross-modal prompt injections or credential exfiltration logic.

I recently built scankii, an open-source static analyzer purpose-built for AI agent skills. It uses a dual-engine (AST + Semantic LLM) approach to catch these exact vulnerabilities before runtime.

Proposal:
I would love to open a PR to add a simple .github/workflows/scankii-scan.yaml file to this repository. It will automatically run scankii against any new skills submitted via Pull Request, ensuring that the skills hosted here remain 100% safe for developers to download and execute in their agent environments.

Let me know if you are open to this, and I'll submit the PR!