The Agent Skills Directory

[SAFE]: The skill implements strict configuration boundaries in its YAML frontmatter, specifically disallowing tools like curl, wget, nmap, and rm that could be used for unauthorized network access or file system modification. It correctly limits its scope to reading files and executing a Python-based validation script.
[SAFE]: The Python script scripts/define_scope.py performs localized parsing of YAML data and IP address validation using the standard ipaddress library. It does not contain any remote code execution, persistence, or data exfiltration logic.
[SAFE]: While the skill references external documentation for cloud provider IP ranges (AWS, GCP, Azure), it does not automatically fetch them at runtime, avoiding risks associated with unverified remote downloads. The included illustrative ranges are hardcoded as static data within the script.
[SAFE]: The skill includes explicit security logic to prevent legal and operational issues, such as detecting when in-scope targets overlap with out-of-scope exclusions or third-party infrastructure.

defining-pentest-scope