Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →superset-sh/doctor
Diagnose and fix Superset problems — connection failures, offline hosts, terminals not attaching, auth or update issues.
Passed 19 of 24 test cases with the skill, and 5 of 24 without it.
and 18 more in the full report
14 fixed5 passed either way5 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 14 test cases the skill fixed. Same model, same prompt, with and without the skill.
I just ran `superset start` to bring up a stopped service that previously caused `superset deploy` to fail. What is the mandatory next step to complete the diagnostic workflow?
| Checked | Without skill | With skill |
|---|---|---|
| Re-runs the originally failing command after applying the fix. | Passed | Passed |
| Explains both the cause and remediation details upon successful verification. | Missed | Passed |
The instructions your model reads when the skill loads.
Diagnose first, change one thing at a time, verify after each change.
superset status — host service healthsuperset auth whoami — auth + active orgsuperset hosts list — host reachabilitysuperset --version| Signature | Fix |
|---|---|
whoami fails / session expired | superset auth login |
| Host shows offline | superset hosts wake <id>; confirm the machine is awake and online |
| Host service not running | superset start |
| CLI and desktop app version mismatch, or stale CLI | superset update |
| App-side misbehavior (macOS) | read the newest entries in ~/Library/Logs/Superset/main.log for errors |
Propose the matching fix and get the user's go-ahead before running anything that changes state. Never delete data as a "fix".
Re-run the originally failing action. If it works, say exactly what was wrong and what fixed it.
If unresolved, offer to file it via the superset:feedback skill — carry over the collected diagnostics (versions, status output, the relevant log excerpt) so the report arrives pre-triaged. Ask before including any log content; logs can contain paths and project names.
test cases passed, vs 5 / 24 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.