See what people built during the Apps HackathonExplore the gallery

Thread

J
Janakiraman3:27 PMOpen in Slack
Hi team,
I'm getting an error while installing microsoft__playwright-mcp. Can anyone help me with a solution?

8 replies
A
Archestra3:27 PMOpen in Slack
Hey Janakiraman! The error "HTTP protocol is not allowed whe..." suggests Playwright MCP is trying to use HTTP transport, which isn't supported in hosted mode. *Quick fix: Click "Edit config" and make sure the server is configured with stdio* transport (command + args), not HTTP/SSE. Example config: ``json { "command": "npx", "args": ["@playwright/mcp@latest"] } `` If you're still stuck, click "View logs" and share the full error β€” happy to dig deeper! πŸ€– Slack Agent
J
Janakiraman3:28 PMOpen in Slack
Log

No pod found for this deployment.

--- Kubernetes Events ---
Failed to retrieve deployment events```
A
Archestra3:28 PMOpen in Slack
That "No pod found" + "HTTP protocol not allowed" combo means the server config is still set to HTTP/SSE transport instead of stdio. Click *Edit config and replace whatever's there with this: ``json { "command": "npx", "args": ["@playwright/mcp@latest"] } ` Make sure there's no url, transport, or sse field β€” just command and args`. Then hit Reinstall*. πŸ€– Slack Agent
J
Janakiraman3:30 PMOpen in Slack
tried but still getting error
A
Archestra3:30 PMOpen in Slack
Can you click *Edit config* and paste the full JSON here? That'll show whether there's still an HTTP/SSE field sneaking in. πŸ€– Slack Agent
MK
Matvey Kukuy (archestra team)4:39 PMOpen in Slack
Hey @user seems like it's some issue with k8s deploy if mcp servers on your cluster, do you see mcp pods running in your k8s?
J
Janakiraman3:27 PMOpen in Slack
Hi @user
I'm getting error message like "Unable to connect to the server: dial tcp: lookup 48E7E20ACF70BDA074B63F00911920B9.gr7.us-east-1.eks.amazonaws.com: no such host"
A
Archestra3:27 PMOpen in Slack
That's an EKS cluster DNS resolution error β€” your Archestra instance can't reach the Kubernetes API server. This is a cluster connectivity issue, not an MCP config issue. Since Matvey is already helping you debug the k8s setup, I'll let him continue β€” he's got the context on your cluster config. πŸ€– Slack Agent