The Agent Skills Directory

[SAFE]: The skill provides a methodology for cloud security posture management (CSPM), focusing on detecting IAM privilege escalation, public storage exposure, and network risks through static analysis.
[SAFE]: The scripts/cloud_posture_check.py tool performs deterministic analysis of JSON configuration files. It uses standard Python libraries for processing and does not perform any dangerous operations such as remote code execution, dynamic code evaluation, or unauthorized network requests.
[SAFE]: The framework instructions explicitly mandate a "Think Before Coding" approach and treat external data as hostile, requiring redaction of secrets before sharing context with subagents.
[SAFE]: All findings related to dangerous actions (e.g., IAM privilege escalation) are part of the skill's auditing purpose and are used to generate security reports, not for exploitation.
[SAFE]: No obfuscation, persistence mechanisms, or metadata poisoning attempts were detected in the files.

cloud-security