The Agent Skills Directory

[PROMPT_INJECTION]: The skill is designed to review and analyze existing SKILL.md files, which introduces a surface for indirect prompt injection where malicious instructions in a reviewed file could attempt to influence the agent.
Ingestion points: SKILL.md files and directory structures provided for review or packaging, which are processed in scripts/quick_validate.py.
Boundary markers: Absent. The instructions do not define explicit delimiters or warnings to ignore instructions found within the files being analyzed.
Capability inventory: File system write operations and directory creation in scripts/init_skill.py, and zip archive creation in scripts/package_skill.py.
Sanitization: Present. The scripts/quick_validate.py utility uses yaml.safe_load() to securely parse frontmatter and includes validation for the length and character content of metadata fields.
[COMMAND_EXECUTION]: The scripts/init_skill.py script creates new Python files and uses chmod(0o755) to make them executable. This is a standard part of the tool's scaffolding functionality but involves the programmatic creation of executable content.

skill-creator-plus