The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes several utility scripts (e.g., init_skill.py, export_skill.sh) that perform legitimate file system operations, such as creating directories, copying templates, and compressing files into ZIP archives using standard system utilities like zip and mkdir.
[EXTERNAL_DOWNLOADS]: The distribution scripts (publish_skill.sh and export_skill.sh) perform network operations to upload packaged skills to the system's backend API (TASK_API_DOMAIN). These operations are authenticated and directed toward the vendor's own infrastructure as part of the intended workflow.
[CREDENTIALS_UNSAFE]: Authentication is handled via an environment-provided auth_token from the TASK_INFO variable. This token is used dynamically for API requests and is not hardcoded within the source code.
[PROMPT_INJECTION]: The skill provides instructions and examples for creating well-structured prompts but does not contain any patterns attempting to bypass safety filters or override system instructions.

skill-creator