The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run and subprocess.Popen in auto_fix.py, watch.py, and reporting.py to manage the local pipeline. These calls trigger internal Python modules (e.g., sparta.pipeline_duckdb.12_qra) and system utilities like grep, tail, and pgrep for log monitoring and data maintenance within the project environment.
[EXTERNAL_DOWNLOADS]: The adversarial.py module uses the httpx library to fetch technique documentation from MITRE ATT&CK (attack.mitre.org). This is performed to verify that the local technique content matches the authoritative remote source. MITRE is a well-known and trusted service in the cybersecurity domain.
[SAFE]: The skill implements a 'Brandon Bailey' domain expert persona to perform semantic validation of space cybersecurity terminology. This is used as a quality control heuristic to ensure generated content contains appropriate technical jargon (e.g., SATCOM, TT&C, Ground Segment) and is not a behavioral override or safety bypass mechanism.

reality-check-sparta