The Agent Skills Directory

[SAFE]: The skill content is focused on professional security advisory services and does not contain malicious instructions or safety bypass attempts.
[COMMAND_EXECUTION]: The skill includes instructions to run three local Python utility scripts: compliance_tracker.py, risk_register_manager.py, and security_posture_scorer.py. These scripts are used for processing local security data and generating reports. Analysis of the source code confirms they only use standard Python libraries (argparse, json, sys, datetime) and do not perform dangerous operations like arbitrary command execution or system modification.
[DATA_EXPOSURE]: No hardcoded credentials, secrets, or sensitive file paths were identified. The scripts read data from user-provided JSON files or use internal defaults without performing any network exfiltration.
[REMOTE_CODE_EXECUTION]: The skill does not download or execute code from external sources. All executable logic is contained within the provided script files.
[PROMPT_INJECTION]: The instructions in SKILL.md are professional and context-aligned. No attempts to override agent behavior or bypass safety filters were found.

ciso-advisor