The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to run the provided Python scripts scripts/init_skill.py and scripts/package_skill.py during the skill creation workflow.
[COMMAND_EXECUTION]: The scripts/init_skill.py script dynamically creates a new Python file (example.py) and modifies its file permissions using chmod(0o755) to make it executable.
[PROMPT_INJECTION]: The skill contains a vulnerability surface for indirect prompt injection as it ingests untrusted user input (descriptions, scenarios, and task examples) to generate the body of a new SKILL.md file and associated script templates.
Ingestion points: User-provided descriptions of skill functionality in Step 1 and content updates in Step 4.
Boundary markers: None identified. The instructions do not define delimiters or provide warnings to the agent to treat user-supplied examples as untrusted data.
Capability inventory: The agent is granted the capability to write files to the local file system and execute the provided helper scripts.
Sanitization: None identified. User input is directly interpolated into file templates without escaping or validation.

skill-creator