Superset 사용 현황을 살펴보고 아직 쓰지 않는 기능 중 업무 방식에 맞는 항목을 제안합니다. 각 기능을 설명하고, 사용자가 고른 예약 실행 에이전트나 병렬 작업 공간, 공유 작업 등을 설정합니다.
- 사용하지 않는 Superset 기능 찾기
- 에이전트 예약 실행 설정
- 여러 작업 공간에서 병렬 작업
Superset의 버그, 기능 요청, 일반 의견을 정리해 제출합니다. 공식 문서를 먼저 확인하고 보고서 초안을 만든 뒤, 사용자가 내용을 검토하고 선택한 비공개 채널이나 공개 GitHub Issue로 보냅니다.
Collect and submit feedback about Paradev (formerly Superset) (bug reports, feature requests, or general feedback) privately to the Paradev team or as a public GitHub issue. Use when the user wants to report a Paradev bug, request a feature, or send feedback about Paradev.
실행 점수 미검증
설치 방법 보기이 스킬은 Superset의 버그, 기능 요청, 일반 의견을 팀이 검토하기 쉬운 보고서로 정리합니다. 먼저 공식 문서에서 기존 해결 방법이나 이미 제공되는 기능이 있는지 확인합니다. 보고가 필요하면 재현 절차, 기대한 동작, 실제 동작 등의 정보를 모아 초안을 보여 줍니다. 사용자는 비공개 채널로 보낼지 공개 GitHub Issue로 작성할지 선택합니다.
버그를 명확히 설명하기 어렵거나, 원하는 기능과 그 이유를 간결하게 전달하고 싶을 때, 팀에 비공개로 의견을 보내고 싶을 때 사용할 수 있습니다. 가능한 경우 버전과 실행 환경도 확인합니다. 공개 Issue를 만들기 전에는 중복 보고를 찾아보고, 기존 보고가 있으면 그곳에 댓글을 남기는 방법도 제안합니다.
비공개 제출에는 superset CLI와 로그인이 필요합니다. 사용할 수 없으면 이메일에 넣을 문안을 준비합니다. 공개 제출에는 gh 또는 브라우저를 사용합니다. 초안을 확인한 뒤 사용자가 명시적으로 선택해야 전송합니다. 로그나 진단 정보는 동의를 얻은 경우에만 첨부합니다. 같은 문제에 관한 추가 정보는 비공개 보고 확인 이메일에 답장으로 보내야 합니다.
이 소개문은 공개된 SKILL.md를 바탕으로 AI(OpenAI Codex)가 작성했습니다. 정확한 사양은 아래 원문을 확인하세요.
AI가 SKILL.md의 내용에서 주의가 필요한 작업과 도입 전제를 확인했습니다. 점수는 실제로 실행한 결과(실행 검증)만으로 매깁니다.
점수에는 넣지 않았습니다. 2026년 10월 29일 기준으로 AI(OpenAI Codex)가 판정했습니다. 참고용이므로 설치 전에 원문도 확인해 주세요.
설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.
paradev is an alias for the superset CLI: every superset command here also runs as paradev.
Check Paradev's documentation for an existing answer first. When feedback still needs reporting, turn it into a short, scannable report and submit it where the user chooses. Treat whatever they wrote after the command as the seed.
The reader is a Paradev engineer triaging dozens of reports. They should get the point from the title, the full picture from the summary bullets, and only read further when they need detail. Cut everything that doesn't help them reproduce or decide.
Every private submission CCs the reporter, so they already have a copy in their inbox with a subject starting [Feedback: ...]. A reply to that email lands on the same thread; a new submission opens a new one and the team has to reconcile them by hand.
If the seed reads like a continuation ("follow-up", "again", "still happening", "update on", "retraction", "root cause", "I already sent", "third time"), or the user filed something on this topic earlier in the session, ask whether they already sent a report about it. If they did, tell them to reply to that email with the new information and stop here. Only file fresh when they confirm it's a different issue.
Never title a report "Follow-up:", "ROOT CAUSE:", or "RETRACTION"; those belong in a reply.
Search the Paradev documentation for the user's question, symptom, or requested capability. Use docs_search with relevant keywords, then docs_read on promising results when the Paradev docs MCP tools are available. Otherwise fetch https://docs.paradev.com/llms.txt and read the relevant linked pages, or use web search restricted to docs.paradev.com. Use generic product terms and error messages; keep private paths, account details, and logs out of search queries.
Read the actual page before treating a result as an answer. Check that it applies to the user's platform, version, and circumstances. If it does, answer directly with the relevant steps and a link to the source. For a requested feature that already exists, explain how to use it. For a bug, offer the documented fix and verify the outcome before treating it as resolved. A related article or workaround alone does not mean the issue is resolved.
If the answer resolves the request and the user has not asked to file anyway, stop without drafting a report. If it does not help, or the user still wants to send feedback, continue below and include the relevant article and what was tried in the draft. Do not make them repeat troubleshooting already covered in the conversation. If the docs are unavailable or no answer matches, continue with the report; do not block submission or claim no answer exists when the lookup failed.
Run in parallel; skip anything that fails:
superset --version and uname -sm for the environment linesuperset auth whoami for the signed-in user/org (private submissions only)superset status --json for the host service; its hostServiceVersion is the desktop app's version when the app spawned the serviceDon't include repository contents, terminal output, or logs unless the user explicitly agrees when asked; they can contain private paths and code.
Many reports describe bugs already fixed in a newer release. Desktop, CLI, and host service share one version number, so compare what's running against the latest release:
superset update --check --json (standalone CLI). If it says the CLI is bundled with the desktop app, the CLI version is the app version; get the latest from gh release download cli-latest -R paradevai/paradev -p version.txt -O - instead.hostServiceVersion from superset status --json if it differs.If anything is behind, say so before drafting and ask whether they want to update first (the desktop app updates itself from its menu; a standalone CLI with superset update). If they'd rather file now, note the versions in the report.
Symptoms like "host not registered", "hosts list is empty", "automations create fails", "host shows offline", "terminal won't attach", or "workspace missing from the sidebar" are the most common reports and usually have a known cause. Before drafting, run the doctor skill's snapshot (superset status, superset auth whoami, superset hosts list) and apply its signature table. If that fixes it, there's nothing to report. If it doesn't, keep the superset status --json output for the report; it's what the team asks for first.
--diagnostics bundle: versions, OS, and the last 200 lines of the app log and the host-service log. Tell the user logs can contain file paths and project names before they agree. Prefer the bundle over attaching raw log files; attach a file only when the tail isn't enough. Attachments must total under about 3 MB; a single larger log is cut to its tail automatically.Classify as bug, feature request, or general feedback from their words; ask only if genuinely ambiguous.
One report per root issue, one request per ask. Two symptoms of the same failure are one bug. Three unrelated wishes are three feature requests; offer to split them, then file each separately. Don't file a second report for a different symptom of something the user just filed.
<Surface>: <symptom or ask>, under 72 characters, no trailing period.
Good: Terminal: pane goes blank after waking from sleep, Sidebar: let me pin an automation to the top.
Bad: Bug with terminal, Terminal rendering broken because xterm loses WebGL context on resume.
Plain text that also reads well as markdown: section labels on their own line, - bullets, numbered steps. No headings, bold, or tables; the private path is delivered as a plain-text email. Keep the user's own words where they're precise, tighten where they ramble. A one-line seed still gets Summary bullets.
Every report starts with a Summary of 2-4 bullets that stand alone. Someone reading only those bullets should know what's wrong (or wanted), how bad it is, and how often it happens. Summary bullets are observations, never hypotheses.
Bug:
Summary
- Terminal pane goes blank after the Mac wakes from sleep
- Happens every time; reloading the window fixes it
- Started in 1.21.0, didn't happen in 1.20.x
Steps to reproduce
1. Open a workspace with one terminal running
2. Close the lid for a minute, then open it
3. Click into the terminal
Expected
The terminal repaints and keeps its scrollback.
Actual
The pane is solid black until the window is reloaded.
Investigation
Hypothesis: main.log shows "WebGL context lost" at the moment of wake, so the renderer may not be reacquiring it.
Host status
{"running":true,"healthy":true,"cloudRegistered":false,...}
Environment
Paradev 1.21.0, macOS 26.0 arm64
Hypothesis:. Never promote them into the title or Summary; reporters have been confidently wrong, and a wrong cause in the title sends triage down the wrong path.superset status --json output, only for the bug classes in step 1.Feature request:
Summary
- I want to pin an automation so it stays at the top of the sidebar
- I check the same two automations many times a day and scroll past twenty others to find them
Today
Automations are sorted by last run, so the ones I care about move around.
Proposal
A pin action in the automation's context menu, pinned items listed first.
General feedback: Summary bullets, then one short paragraph of detail if there's more to say.
Omit any section with nothing to say. Steps, Expected, Actual, Investigation, Host status, and Environment are for bugs only. Target under 150 words; go longer only when the extra words help reproduce.
Show the full draft, then ask the user (use the ask_user tool if available, otherwise a plain question) with exactly these options:
Never submit anything before the user explicitly picks 1 or 2. Loop on edits.
Private path:
superset feedback --help >/dev/null 2>&1 exits 0 (the exit code alone says the command exists; it works on every CLI version and needs no jq), submit via stdin (note: --body-file=- with the equals sign; a space-separated - is rejected by the parser):
superset feedback submit --type <bug|feature|general> --title "..." --body-file=- <<'EOF'
<drafted report>
EOF
Only when the user agreed to them in step 1, add --attach /path/to/screenshot.png (comma-separated paths, about 3 MB total) and/or --diagnostics. The submission is sent from the user's Paradev account, a copy is CC'd to them, and the team replies to their account email.superset auth whoami fails), offer superset auth login first; if declined, fall back to email: give the user a clickable mailto link (mailto:support@paradev.com?subject=<url-encoded title>&body=<url-encoded body>) and also print the raw draft so they can copy it.Public path:
gh search issues -R paradevai/paradev "<key terms>" --limit 5. If an existing issue matches, show it and offer to comment there (gh issue comment) instead of opening a new one; only create a fresh issue if the user confirms it's genuinely different.gh is installed and gh auth status succeeds: gh issue create -R paradevai/paradev --title "..." --body "..." (write the body via a heredoc or temp file, never inline-escape).https://github.com/paradevai/paradev/issues/new?title=<url-encoded>&body=<url-encoded>.Report back the issue URL (public) or a confirmation of what was sent and to whom (private). Remind them that more information on the same issue goes as a reply to the CC'd email, not a new report. If anything failed, show the draft so the user's writing is never lost.
아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.
로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.
아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.
Superset 사용 현황을 살펴보고 아직 쓰지 않는 기능 중 업무 방식에 맞는 항목을 제안합니다. 각 기능을 설명하고, 사용자가 고른 예약 실행 에이전트나 병렬 작업 공간, 공유 작업 등을 설정합니다.
반복 작업을 Superset 자동화로 설정합니다. 에이전트 지침, 일정 또는 이벤트 조건, 실행 작업 공간을 정리하고 명시적인 확인을 받은 뒤 자동화를 만듭니다. 저장된 조건과 첫 실행 결과를 살펴 지침을 조정합니다.
Superset 작업 공간의 브라우저에서 페이지를 열고 클릭하거나 입력하며, 화면과 콘솔 정보를 확인합니다. 개발 중인 웹 앱 미리 보기, 폼 동작 및 화면 이동 점검에 활용할 수 있습니다.
Stripe 공식 문서와 API 매개변수 정의를 확인해 결제, 구독, 청구서, 웹훅 연동을 구현합니다. 적절한 제품 선택부터 샌드박스 테스트와 웹훅 처리까지 다루며, 결제와 구독 상태를 웹훅 이벤트로 판단합니다.
Chrome DevTools Protocol로 실행 중인 Electron 데스크톱 앱을 조작해 UI 변경을 확인하고 문제를 재현합니다. 실제 사용 흐름을 따라 화면 전환을 검증하고, 스크린샷과 경로·스크롤 위치·포커스·저장 상태를 기록해 비교합니다.
GitHub Actions 실패 로그에서 최초 오류를 찾아 코드 결함, 환경 문제, 간헐적 실패, 기존 장애로 구분합니다. 기준 브랜치와 PR 변경 사항을 비교해 관련성을 설명하고 다음 조치를 정리합니다.