← back to grpo-rl-training

Skill Scanner · grpo-rl-training

independent scanner by Cisco AI Defense · skill by qcmuu · how it works ↗

FAILmax severity: CRITICAL

Dangerous code execution functions that can execute arbitrary code

scanned 2026-08-22

Findings (1)

CRITICALcommand_injection

Dangerous code execution functions that can execute arbitrary code

examples\reward_functions_library.py:354

What the verdicts mean

Skill Scanner reports on SkillMD's shared five-tier scale. See how Skill Scanner works ↗.

PASS

Reported as safe — no findings

CAUTION

Findings up to MEDIUM severity

WARNING

Findings of HIGH severity

FAILthis skill

Findings of CRITICAL severity

INCONCLUSIVE

Scan could not complete