The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard local commands including mkdir -p, git mv, and git commit to manage repository structures.
[COMMAND_EXECUTION]: It executes pnpm ai-hero-cli internal lint to validate the created exercises, which is a common practice in development environments.
[SAFE]: No malicious patterns such as obfuscation, credential theft, or unauthorized network activity were detected. All instructions are consistent with the skill's curriculum management purpose.
[PROMPT_INJECTION]: The skill ingests user-provided plans (Ingestion point: exercise plan parsing) to perform shell operations (Capabilities: mkdir, git, pnpm) without explicit delimiters or sanitization. This lack of boundary markers represents a surface for indirect prompt injection, though it is assessed as safe within the intended local development context.

zj-scaffold-exercises