.github/workflows/docker-build.yml (1)

299-319: Tail scope-app.log before exiting on startup timeout.

Right now a startup failure turns into a generic 120s timeout, and the actual exception only exists in the artifact. Dumping the last log lines here makes failed runs much faster to diagnose.

Proposed change

       if [ $ATTEMPT -eq $MAX_ATTEMPTS ]; then
         echo "App did not start in time"
+        if [ -f scope-app.log ]; then
+          echo "::group::scope-app.log (tail)"
+          tail -n 200 scope-app.log
+          echo "::endgroup::"
+        fi
         exit 1
       fi

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In @.github/workflows/docker-build.yml around lines 299 - 319, When the startup
loop reaches the timeout (when ATTEMPT equals MAX_ATTEMPTS), print the last
lines of the background process log to aid debugging: update the timeout branch
in the startup script (the while loop and the subsequent if that compares
ATTEMPT and MAX_ATTEMPTS) to echo the timeout message, then run a tail of
scope-app.log (e.g., tail -n 200 scope-app.log) to dump recent logs to the
workflow output before exiting with status 1; keep the existing variables
MAX_ATTEMPTS, ATTEMPT, and the background invocation uv run daydream-scope >
scope-app.log 2>&1 & intact.