| name | coreweave-debug-bundle |
| description | Collect CoreWeave cluster diagnostics for support tickets.
Use when preparing a support case, collecting GPU node status,
or documenting pod failures.
Trigger with phrases like "coreweave debug", "coreweave support",
"coreweave diagnostics", "collect coreweave logs".
|
| allowed-tools | Read, Bash(kubectl:*), Bash(tar:*), Grep |
| version | 1.0.0 |
| license | MIT |
| author | Jeremy Longshore <jeremy@intentsolutions.io> |
| tags | ["saas","gpu-cloud","kubernetes","inference","coreweave"] |
| compatible-with | claude-code |
CoreWeave Debug Bundle
Instructions
#!/bin/bash
set -euo pipefail
BUNDLE="coreweave-debug-$(date +%Y%m%d-%H%M%S)"
mkdir -p "$BUNDLE"
kubectl get nodes -o wide > "$BUNDLE/nodes.txt"
kubectl get pods --all-namespaces -o wide > "$BUNDLE/pods.txt"
kubectl get events --sort-by=.lastTimestamp > "$BUNDLE/events.txt"
kubectl describe nodes | grep -A10 "Allocated resources" > "$BUNDLE/gpu-allocation.txt"
for pod in $(kubectl get pods --field-selector=status.phase=Failed -o name); do
kubectl logs "$pod" --tail=100 > "$BUNDLE/$(basename $pod)-logs.txt" 2>&1
done
tar -czf "$BUNDLE.tar.gz" "$BUNDLE" && rm -rf "$BUNDLE"
echo "Bundle: $BUNDLE.tar.gz"
Resources
Next Steps
For rate limit handling, see coreweave-rate-limits.