The Agent Skills Directory

[SAFE]: The skill performs local file system operations, including directory creation, file templating, and symlinking, which are consistent with its stated purpose of automating skill development and management.
[COMMAND_EXECUTION]: The skill instructs the agent to execute standard shell commands such as git, mkdir, sed, and ln to manage the local development environment. These commands are used transparently for scaffolding, metadata extraction, and installation.
[DATA_EXPOSURE]: The skill accesses local git configuration (user.name, user.email) to populate metadata fields in generated skills, which is standard practice for authorship attribution in development tools and does not involve external transmission.
[REMOTE_CODE_EXECUTION]: No patterns of downloading and executing code from external or untrusted sources were detected. All scripts and templates are contained within the local repository.
[PROMPT_INJECTION]: The skill processes user-provided strings for skill names and descriptions to generate new files. While this creates a surface for indirect prompt injection, the skill includes automated validation steps and operates as a scaffolding tool where user oversight is expected.

skill-creator