The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes platform-specific commands such as write, edit, skills, and install in SKILL.md to manage file operations and skill lifecycles. It also mentions usage of the npx skills CLI tool for skill discovery.
[EXTERNAL_DOWNLOADS]: Instructions in SKILL.md describe importing skill content from the skills.sh marketplace and public GitHub repositories to support workflow reuse.
[INDIRECT_PROMPT_INJECTION]: The skill framework processes data from marketplace imports and user-provided instructions to generate local agent guidance.
Ingestion points: Data is pulled from external sources using the skills({ query }) and import({ source }) functions described in SKILL.md.
Boundary markers: The skill provides optimization guidelines in references/description-optimization.md to ensure that descriptions act as specific triggers rather than generic summaries.
Capability inventory: The skill has the ability to write to project and global directories and modify file contents.
Sanitization: SKILL.md contains an explicit security reminder to only include executable scripts in the scripts/ directory that have been thoroughly reviewed for safety.

open-knowledge-write-skill