The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest user requirements to generate new skill structures, instructions, and scripts. This creates a surface area where malicious user input could influence the generated skill's behavior or logic.
Ingestion points: User descriptions and examples of desired skill functionality (Step 1 and Step 2 in SKILL.md).
Boundary markers: The instructions recommend using imperative form and specific frontmatter, but do not explicitly define delimiters for untrusted user input within the generated files.
Capability inventory: The skill uses scripts/init_skill.py to create file structures and scripts/package_skill.py to bundle them. It also explicitly encourages the agent to write and test new scripts (Step 4).
Sanitization: The skill mentions normalizing user-provided titles to hyphen-case, but lacks comprehensive sanitization for the content of the generated scripts or documentation.
[DYNAMIC_EXECUTION]: The skill workflow involves generating executable scripts (Python/Bash) and requires the agent to run them to ensure they work as expected.
Evidence: Step 4 states, 'Added scripts must be tested by actually running them to ensure there are no bugs'.
Pattern: This follows a script generation and execution pattern, where the agent is prompted to create code and then execute it within its environment.

skill-creator