You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Streamable http MCP server handshaking failed #5208
MCP client for custom failed to start: handshaking with MCP server failed: Send message error Transport
[rmcp::transport::worker::WorkerTransport<rmcp::transport::streamable_http_client::StreamableHttpClientWorker<reqwest::a
sync_impl::client::Client>>] error: Transport channel closed, when send initialized notification
Can you reproduce this issue with any other MCP servers? Does your MCP server support dynamic client registration? Does it require anything else like custom headers?
Can you reproduce this issue with any other MCP servers? Does your MCP server support dynamic client registration? Does it require anything else like custom headers?
@gpeal Yes, I found that others have similar feedback. We use different MCP servers. To make it easier for you to test, I provide you with an MCP link with an authorization code. In cursor or claudecode, you can directly add the URL to obtain the tool list and call it. It will only report an error in codex.
I updated codex to v0.47.0 but unfortunately it did not help. @moxi000 your server gives me
MCP client for `allflow` failed to start: handshaking with MCP server failed: Send message error Transport
[rmcp::transport::worker::WorkerTransport<rmcp::transport::streamable_http_client::StreamableHttpClientWorker<reqwest::async
_impl::client::Client>>] error: Transport channel closed, when send initialized notification
Reacted by David Vincent Fischer and Serhii Khalymon
MCP client for langchain failed to start: handshaking with MCP server failed: Send message error Transport
[rmcp::transport::worker::WorkerTransport<rmcp::transport::streamable_http_client::StreamableHttpClientWorkerreqwest::async_impl::client::Client>] error: Client error: HTTP status server error (500 Internal
Server Error) for url (https://docs.langchain.com/mcp), when send initialize request
config.toml:
Is this a related issue? Have I done something wrong?
MCP client for `langchain` failed to start: handshaking with MCP server failed: Send message error Transport [rmcp::transport::worker::WorkerTransport] error: Client error: HTTP status server error (500 Internal Server Error) for url (https://docs.langchain.com/mcp), when send initialize request
config.toml:
Is this a related issue? Have I done something wrong?
Sounds like there was an issue on their end, I just tried and it worked
I had the same issue. Fixed by doing the following:
codex mcp remove linear
codex mcp add --url https://mcp.linear.app/mcp linear
codex mcp login linear -c experimental_use_rmcp_client=true
# Confirm it's working
codex mcp get linear
I had the same issue. Fixed by doing the following:
codex mcp remove linear
codex mcp add --url https://mcp.linear.app/mcp linear
codex mcp login linear -c experimental_use_rmcp_client=true
# Confirm it's working
codex mcp get linear
Activity
github-actions commentedon Oct 15, 2025
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
moxi000 commentedon Oct 17, 2025
same problem, however correctly work in cursor and claudecode
gpeal commentedon Oct 17, 2025
Can you reproduce this issue with any other MCP servers? Does your MCP server support dynamic client registration? Does it require anything else like custom headers?
moxi000 commentedon Oct 18, 2025
@gpeal Yes, I found that others have similar feedback. We use different MCP servers. To make it easier for you to test, I provide you with an MCP link with an authorization code. In cursor or claudecode, you can directly add the URL to obtain the tool list and call it. It will only report an error in codex.
https://mcp.allflow.cn/message?token=uwP1eH5SoKaA2kRGGKjk9GCbbkvtk06RUVUtNbZH
hwait commentedon Oct 18, 2025
I can confirm the issue with Linear:
It worked well couple of days ago.
moxi000 commentedon Oct 18, 2025
I still can’t use it normally.
hwait commentedon Oct 18, 2025
I updated codex to v0.47.0 but unfortunately it did not help.
@moxi000 your server gives me
dominicdill commentedon Oct 22, 2025
MCP client for
langchainfailed to start: handshaking with MCP server failed: Send message error Transport[rmcp::transport::worker::WorkerTransport<rmcp::transport::streamable_http_client::StreamableHttpClientWorkerreqwest::async_impl::client::Client>] error: Client error: HTTP status server error (500 Internal
Server Error) for url (https://docs.langchain.com/mcp), when send initialize request
config.toml:
Is this a related issue? Have I done something wrong?
gpeal commentedon Oct 22, 2025
Sounds like there was an issue on their end, I just tried and it worked
moxi000 commentedon Oct 22, 2025
@gpeal Can you try with my link?https://mcp.allflow.cn/message?token=uwP1eH5SoKaA2kRGGKjk9GCbbkvtk06RUVUtNbZH
rafaelquintanilha commentedon Oct 22, 2025
I had the same issue. Fixed by doing the following:
gpeal commentedon Oct 22, 2025
I do see the issue with that mcp server. I'll investigate separately. Feel free to take a look as well. It's all open source.
gpeal commentedon Oct 22, 2025
@moxi000 your issue will be resolved by modelcontextprotocol/rust-sdk#497
moxi000 commentedon Oct 23, 2025
Thanks
hwait commentedon Oct 23, 2025
Thank you, Rafael, worked well!
gpeal commentedon Oct 23, 2025
I'm going to close this out now because the allflow issue was fixed by #5542
Please open new issues for anything that happens in 0.48 or later
[MCP] Update rmcp to 0.8.3 (#5542)
[MCP] Update rmcp to 0.8.3 (openai#5542)
[MCP] Update rmcp to 0.8.3 (openai#5542)
[MCP] Update rmcp to 0.8.3 (openai#5542)
[MCP] Update rmcp to 0.8.3 (openai#5542)