根据课程目标生成教案、互动题、作业与分层教学建议。；use for education, lesson-plan, teaching workflows；do not use for 生成违规内容, 替代教师现场判断.

This skill appears to do what it says: produce lesson-plan drafts from local inputs using a bundled Python script. Before running it, inspect scripts/run.py (it is included) and only pass input paths you trust — the script will read and sample files from any directory you point it to. Do not run it against system/root or sensitive directories, and run in a sandbox or dedicated workspace if you want to be extra cautious. Confirm there are no expectations of network access (none are present) and verify outputs before using them in production or publishing.

Type: OpenClaw Skill Name: classroom-lesson-pack Version: 1.0.0 The bundle contains a script (scripts/run.py) that includes extensive file-scanning and auditing logic entirely unrelated to its stated purpose of generating classroom lesson plans. The script features a 'pattern_report' function with regex patterns designed to identify sensitive data such as API keys, tokens, and private URLs, as well as 'directory_audit' and 'skill_audit' modes. While the default configuration in resources/spec.json is set to a benign text-formatting mode, the presence of these latent capabilities for searching the filesystem for secrets and dangerous commands is highly irregular for an educational tool and represents a significant risk if the agent is redirected to use these functions.