The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes bunx to execute the @exercode/problem-utils package. This involves fetching and running a specific package from the npm registry at runtime to validate generated course content.
[COMMAND_EXECUTION]: The skill runs the exercode-problem CLI tool (via bunx) to perform semantic and schema validation on generated materials. The command bunx exercode-problem validate-course <courseDir> is executed in the environment to ensure material integrity.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user input such as course topics and audience descriptions to generate markdown and YAML files. While this creates an ingestion surface for indirect prompt injection, the risk is mitigated by the skill's use of strict YAML schemas, clear markdown boundaries, and a post-generation validation step that checks for structural and semantic errors.

generate-course-materials