enhance: remove auto delete and see logs of mcp failure on setup

[ivyjeong13]

Addresses #4182

[ivyjeong13]

Wanted to test this again before merging, but looks like /logs is no longer working for me when trying to run with the Tavily example case

After getting the 503 from /launch, trying to hit /logs gives me:
500 server running, but readiness check failed: server readiness check failed: MCP server is not healthy

Before, /logs would be successful and I could get the logs info showing that I am missing an environment variable. Example of what was happening before: https://www.loom.com/share/23d35958db214d99a48b99349b42e5e4

Wondering if there was a change with Tavily or on our side and holding off merge to see if I can try with another MCP server!

[ivyjeong13]

Wanted to test this again before merging, but looks like /logs is no longer working for me when trying to run with the Tavily example case

After getting the 503 from /launch, trying to hit /logs gives me: 500 server running, but readiness check failed: server readiness check failed: MCP server is not healthy

Before, /logs would be successful and I could get the logs info showing that I am missing an environment variable. Example of what was happening before: https://www.loom.com/share/23d35958db214d99a48b99349b42e5e4

Wondering if there was a change with Tavily or on our side and holding off merge to see if I can try with another MCP server!

Was able to determine there was a backend change that made it so /logs will return a 500 if the deployment fails for any reason. Waiting on a backend PR to merge #4530 to verify that /logs can be hit if due to invalid/not provided environment variable.

[g-linville]

Backend lgtm