cheatengine-mcp-bridge

SUSPICIOUS. The skill's dangerous capabilities largely match its stated reverse-engineering purpose, so this is not disguised credential theft, but it grants an AI agent highly invasive offensive tooling: arbitrary memory access, debugging, code injection, DLL injection, optional shell execution, and anti-detection guidance. Install sources are mostly standard rather than blatantly malicious, yet the scope is high-risk for an agent skill and not proportionate for normal developer assistance.