The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to run shell commands including mkdir, git commit, git mv, and pnpm ai-hero-cli internal lint to manage and validate exercise directory structures.
[PROMPT_INJECTION]: The skill ingests a user-provided plan to automate directory creation, creating an entry point for untrusted data (Indirect Prompt Injection).
Ingestion points: The agent parses user input in the 'Parse the plan' step in SKILL.md.
Boundary markers: The instructions do not define boundary markers to separate the plan content from the agent's instructions.
Capability inventory: The agent can perform file writes, directory creation, and shell command execution.
Sanitization: There are no instructions for sanitizing user-provided names before they are used in shell commands or file operations.

scaffold-exercises