multi-lens

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The prompt includes an instruction that agent instances run with mode: "bypassPermissions", which is a hidden/deceptive directive to override normal permission constraints and changes agent behavior beyond the skill's stated analytical purpose.

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

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The prompt explicitly instructs spawned agents to run with mode: "bypassPermissions" (i.e., bypass security mechanisms) and to overwrite files on the host, which directs the agent to circumvent permissions and modify machine state.