How to determine the ’target tool‘？

[zggg1p]

Excuse me, how do you determine the ‘target tools’ when designing a malicious response to an injected tool? Do you assume that the attacker can know what tools the platform has?

Looking forward to your answer.

[zrpxx]

Sorry for the delay in response.

The target tools are pre-designated by the attacker. This is related to the attacker's goal: to ensure that the query calls the target tool when it can be retrieved, or rejects the service when it cannot. The attacker does not necessarily need to know what tools the platform has, but they do need a specific tool as their intended target for the attack.