Can we implement an automated analysis tool for a WASM contains mostly open source code and tiny of unknown code?

Imagine that we are given an unknown WASM file and know that it simply calls some open source crates we can easily found on, with probably less than 50 lines in source code of its own logic. The almost completely preserved symbolic information in the function names confirms this, and we can even learn the version of rustc in the original WASM file. Now our task is to restore its own 50-lines logic as much as possible.
Apart from manually analyzing the call chain based on the symbolic information, perhaps we can implement an automated analysis tool? It is certainly possible in theory, since we have enough known information. But the implementation could be complex and perhaps require someone knows rustc and LLVM very well.

I'm now writing a small program that simplifies the AST to just function names and call chains, at least not getting lost in a sea of thousands of lines of stack operations.