stacklok / toolhive-studio / 24345950005 / 1

70%

main: 70%

push

github

feat: show dedicated error modal when another ToolHive server is already running (#1979)

* feat: show dedicated error modal when another ToolHive server is already running

Detect the "already running" error from thv stderr and display a specific
error screen instead of the generic "Oops, something went wrong" modal.

Closes #1977

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* fix: re-fetch toolhive status in health check to fix race condition

The status was fetched once before the health check and passed as a
parameter. By the time the health check failed, the process had already
exited and set processError, but the stale snapshot missed it.

Now checkHealth re-fetches the status in its catch block, correctly
picking up errors like ALREADY_RUNNING. Also updates the error modal
text to mention "ToolHive HTTP server (thv serve)".

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* refactor: reorder guards to check health before registry auth

Health check is more fundamental — if the server is down, registry
auth redirect is wasted work. Moving checkHealth before
handleRegistryAuthRedirect ensures we short-circuit early with
the appropriate error modal.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* test: add tests and HealthCheckErrorCause type for already-running error

Address Copilot review comments:
- Add HealthCheckErrorCause type for type-safe error cause handling
- Add RootErrorComponent test for ALREADY_RUNNING branch
- Add toolhive-manager tests for stderr-driven processError detection
- Simplify cause checks in RootErrorComponent using typed cause

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

3493 of 5801 branches covered (60.21%)

5601 of 8522 relevant lines covered (65.72%)

121.39 hits per line