The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill includes Python scripts (init_skill.py, package_skill.py) designed for local file structure generation and packaging. These scripts use the pathlib module safely and enforce a strict naming convention ([a-z0-9-]+) for user-provided input, effectively mitigating risks of directory traversal or shell injection.\n- [DATA_EXFILTRATION] (SAFE): No code was found that attempts to read sensitive files (e.g., SSH keys, credentials) or transmit data to external network endpoints. All operations are local to the skill's directory.\n- [EXTERNAL_DOWNLOADS] (SAFE): The skill does not download external scripts or execute remote code. It relies on standard library components and a common, safe YAML parser.\n- [PROMPT_INJECTION] (SAFE): The instructions in SKILL.md are purely technical and instructional regarding the creation of skills. There are no attempts to override system instructions or bypass security filters.

skill-creator