The Agent Skills Directory

[SAFE]: The skill serves as a meta-tool for skill development and auditing. A thorough review of the included scripts and instructions revealed no malicious behavior.
[COMMAND_EXECUTION]: The skill utilizes Bash(python:*) to execute local Python scripts included in the package. These scripts (validate_skill.py, check_self_contained.py, test_activation.py) are used for static analysis and do not perform any dangerous operations.
[PROMPT_INJECTION]: The skill's instructions are focused on guiding the creation of expert-level skills and do not attempt to manipulate or bypass the agent's core safety protocols.
[DATA_EXFILTRATION]: There is no evidence of network communication or unauthorized data access. All file operations are confined to local skill development directories.
[EXTERNAL_DOWNLOADS]: The skill mentions standard package installation (e.g., via pip) in its examples and documentation. These refer to trusted libraries such as transformers, pillow, and torch from official registries.

skill-coach