Some usage suggestions

[huihui12021505]

I recently set up an MCP service for a project and tried A.I.G's MCP scanning function.

The whole process was much simpler than I expected: I logged into AIG's web interface at localhost:8088. Since the code is stored on an internal GitLab server, I selected the "Source Code Scan" mode and pasted the path to my local cloned repository. Then the scan began.

The final report revealed a command injection vulnerability that could occur when a tool processes user input. The attack path shown in the report illustrates how attackers could exploit this vulnerability to execute arbitrary system commands.

Overall, A.I.G is easy to use, and the vulnerability descriptions and attack examples in the report are very clear and concise.

However, I have the following suggestions that I hope developers will consider adopting:

1. Improve scanning speed or support batch parallel processing tasks.

2. Support custom headers for remote MCP servers.

Thank you for your attention.

[rsinding]

Thank you for sharing your experience and suggestions!
Scanning speed: We’ve improved the scanning speed by 30% in the latest release and are continuing to optimize it.
Custom headers: This feature is still in development, but we’ll provide an update on its release timing soon.
We appreciate your feedback—it helps us improve the project. If you have any other suggestions or questions, feel free to reach out!

[aigsec]

Hi there, your suggestion about adding custom headers support for remote MCP servers. we’re happy to let you know this feature has been officially released in v3.5.0. You can upgrade to the latest version to start using it right away.