The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands via cmd /c, mise, and npx to perform lifecycle operations such as initializing skills, verifying installations, and running validation scripts.
[EXTERNAL_DOWNLOADS]: The skill interacts with the why8023/agent-skills repository on GitHub and the npm registry. These network operations are intrinsic to the skill's purpose of synchronizing code with a remote central repository and managing dependencies.
[DATA_EXFILTRATION]: While the skill performs git push operations to a remote repository, this is directed to the author's own repository (why8023/agent-skills) as part of a documented synchronization workflow, not an unauthorized exfiltration of sensitive data.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface as it processes user-provided skill names and content.
Ingestion points: Skill names and content provided by the user (or the agent on the user's behalf) are written to local files.
Boundary markers: No specific boundary markers or 'ignore' instructions are used for the created content, as the skill functions as an authoring tool.
Capability inventory: The skill can write files, execute Git commands (commit/push), and run the npx skills utility.
Sanitization: The skill assumes the integrity of the authored content and does not implement specific sanitization routines before file writes or Git commits.

skill-authoring-sync