pyre-code-ml-practice

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The repository intentionally executes untrusted submission code via exec(submission_code) (and exposes a /submit API), which enables remote code execution capable of data exfiltration, credential theft, or system compromise if the grading service is run without strong sandboxing/isolation; otherwise no other explicit backdoors or obfuscated malicious payloads are present.