Skill UI

Generates minimal macOS Seatbelt sandbox profiles with allowlist-based resource rules so apps can be isolated by permissions, useful for restricting file, network, and IPC access when defending against high-risk execution or supply chain threats.

A comprehensive guide detailing advanced security threats in Machine Learning and AI systems. Covers model supply chain attacks (RCE via pickle, poisoning), adversarial examples (FGSM, PGD, physical-world), data privacy breaches, and LLM-specific vulnerabilities. Essential for building robust, secure AI pipelines.

This skill audits Model Context Protocol (MCP) servers to detect critical AI security vulnerabilities. It scans for tool poisoning, tool shadowing, rug pulls, SSRF, and unauthenticated exposure. Use it before integrating new tools or during security reviews to ensure agent reliability and prevent supply chain attacks.

A comprehensive tool for identifying poisoned training data, backdoored ML models, and supply-chain compromises. It leverages IBM's Adversarial Robustness Toolbox (ART) for activation clustering and spectral signature analysis, uses Cleanlab for label quality checks, and enforces integrity via weight-hash verification and safetensors. Essential for securing ML pipelines before training, deployment, or during supply chain audits.