ctf-pwn

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). High-risk: this is an explicit, step-by-step exploit development guide that teaches arbitrary read/write, control-flow hijacking, ret2libc/ROP, spawning shells and bypassing protections—techniques readily usable to gain remote code execution, install backdoors, or exfiltrate data.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.80). The prompt explicitly guides the agent to discover and craft exploits (e.g., inject shellcode, call system("/bin/sh"), ret2libc, test exploits locally) which push the agent toward executing arbitrary code and obtaining shells on the host—actions that can compromise machine state—even though it doesn't explicitly ask for sudo or account/config changes.