thrifty-dispatch

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The script explicitly sends local source files, test outputs, and gate failures to an external LLM service and writes/executes model-produced code, creating high-risk data-exfiltration and backdoor/RCE vectors (writes are sandboxed to the workspace but not content-sanitized).