The Agent Skills Directory

[SAFE]: The skill consists primarily of markdown instructions and reference logic for structural problem analysis. No malicious patterns or behavioral overrides were detected.- [COMMAND_EXECUTION]: The skill includes a helper script scripts/quick_validate.py. This script is a benign utility designed to validate the YAML frontmatter of the skill's main entry point. It uses standard file operations and string matching without performing any dangerous shell commands or network requests.- [DATA_EXFILTRATION]: No network operations or sensitive data access patterns were identified. The skill's primary output involves generating local HTML reports based on provided templates, which is a legitimate and intended functionality.- [PROMPT_INJECTION]: The instructions and the openai.yaml configuration include strict behavioral guards. The framework enforces a 'clarification-before-analysis' gate and explicitly prohibits the agent from engaging in roleplay, slogan-writing, or bypassing safety protocols. It also includes mechanisms to prevent 'focus drift' in complex analytical tasks.

maozedong-maoxuan-skill