bug: panic, nil pointer dereference in Stdio readResponses

[xanarry]

Description

I need some help with a panic issue when using the stdio MCP client.
I encountered the following panic

panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x10 pc=0x5c27bd]

goroutine 27837 [running]:
bufio.(*Reader).ReadSlice(0x0, 0xa)
        /usr/local/go/src/bufio/bufio.go:355 +0x3d
bufio.(*Reader).collectFragments(0x0, 0xa)
        /usr/local/go/src/bufio/bufio.go:455 +0x70
bufio.(*Reader).ReadString(0x0?, 0x0?)
        /usr/local/go/src/bufio/bufio.go:502 +0x1f
github.com/mark3labs/mcp-go/client/transport.(*Stdio).readResponses(0xc000f40240)
        /Users/xanarry/go/pkg/mod/github.com/mark3labs/mcp-go@v0.29.0/client/transport/stdio.go:170 +0x53
github.com/mark3labs/mcp-go/client/transport.(*Stdio).Start.func1()
        /Users/xanarry/go/pkg/mod/github.com/mark3labs/mcp-go@v0.29.0/client/transport/stdio.go:79 +0x2f
created by github.com/mark3labs/mcp-go/client/transport.(*Stdio).Start in goroutine 27955
        /Users/xanarry/go/pkg/mod/github.com/mark3labs/mcp-go@v0.29.0/client/transport/stdio.go:77 +0x8c

my mcp-go version is v0.29.0

I have checked my own code and the MCP client initialization logic, but I couldn't find any obvious reason for c.stdout to be nil at this point. My guess is that this happens when the MCP server subprocess exits unexpectedly, and the client tries to read from a closed or nil pipe.
Should there be a panic recovery or more defensive check in the readResponses goroutine to avoid crashing the whole process?
Or is there any recommended way to handle this scenario gracefully?
Any advice or fix would be appreciated. Thank you!

[xanarry]

I have tried to kill stdio mcp server process all the time, but I can not trigger this pannic, anybody help

[pottekkat]

@xanarry There has been some breaking changes recently but some bugs were also resolved which could be the issue. Could you try the latest release of the SDK? It should not have the issue.

[xanarry]

@pottekkat thanks, I have updated it from v0.29.0 to v0.31.0, and continue to observe whether the problem will recur.

[pottekkat]

Great, thanks. You can reopen this issue if you still have problems.

[xanarry]

@pottekkat hello. pottekkat, I have updated sdk to v0.31.0, but it still quit without any warning signs today, the same panic error