概要
This skill provides a comprehensive auditing framework for AI plugins, ensuring they adhere to function calling best practices and secure coding standards. It automatically detects issues like missing schema constraints, vague tool descriptions, and potential security vulnerabilities like command injection or path traversal across Claude Code, OpenCode, and Codex environments. By applying automated fixes for high-certainty issues, it helps developers build more reliable, efficient, and secure AI-integrated tools while significantly improving the accuracy of model-tool interactions.
