medsci-skills 4.1.0

package/skills/ma-scout/references/project_readme_template.md

@@ -0,0 +1,95 @@
+# MA-Scout Project README Template (PROSPERO-ready)
+
+Load this reference when `/ma-scout` Phase 4 (Folder & README Scaffolding) creates a
+new topic folder. English section headings and PICO/PIRD structure keep the README aligned
+with downstream `/meta-analysis` and `/search-lit` expectations.
+
+This is the English default. For a supervisor / professor-facing Korean README, use
+`project_readme_template_ko.md` (same structure, Korean prompts for the PI-facing fields).
+
+Copy the block below into `{topic_folder}/README.md` and fill in the curly-brace fields.
+
+````markdown
+# {Topic Title} — {MA Type} Meta-analysis ({Supervisor name})
+
+## Overview
+- Supervisor: {Supervisor name} ({affiliation history})
+- Supervisor's area: Pillar {N} — {area name}
+- Status: Planning
+- Priority: rank {N}
+- Created: {YYYY-MM-DD}
+
+## Research Question
+### PICO/PIRD
+- **P**opulation: {specific patient group, disease, setting}
+- **I**ndex test / Intervention: {test or intervention}
+- **C**omparator / Reference standard: {comparator or reference standard}
+- **O**utcome: {DTA: Se/Sp/AUC, Prognostic: HR/OR, adverse events, etc.}
+
+### One-line RQ
+{the complete research question in one sentence}
+
+## Key Gap
+- Existing MAs: {N} ({details, most recent year, scope limitations})
+- Consensus/Scholar Gateway cross-check: {result summary}
+- medRxiv/bioRxiv preprint MA: {yes/no}
+- Registered PROSPERO protocol: {yes (CRD#) / no}
+- {specific gap explanation — why a new MA is needed}
+
+## Professor's Authority
+- {number of related publications, 1-2 representative papers, distinctive contribution}
+- {why this supervisor is well-suited to this topic}
+
+## Preliminary Search ({date})
+### Search Strategy (PubMed)
+```
+{the actual PubMed query used — the E-utilities esearch query verbatim}
+```
+- Total hits: {N} (raw)
+- DTA/outcome extractable (estimated): {N} (×0.15-0.30 discount)
+- Existing MAs: {N} (narrow) / {N} (including broad SRs)
+- Consensus search result: {whether any additional papers were found}
+- bioRxiv/medRxiv: {N} preprints
+
+### Embase Search Strategy (Draft)
+```
+{draft Embase query — including Emtree terms}
+```
+(draft before execution — verify when Embase access is available)
+
+## Target Journal
+| Rank | Journal | IF | MA acceptance rate | Turnaround | Notes |
+|------|---------|-----|--------------------|-----------|-------|
+| 1st | {journal} | {IF} | {high/medium} | {months} | {rationale} |
+| 2nd | {journal} | {IF} | {high/medium} | {months} | {rationale} |
+
+## Timeline (backwards-planned)
+| Stage | Estimated time | Prerequisite |
+|-------|----------------|--------------|
+| Supervisor proposal | {YYYY-MM} | {condition} |
+| PROSPERO registration | +1 week | Supervisor approval |
+| Search complete | +2 weeks | PROSPERO registration |
+| Screening complete | +3 weeks | 2nd reviewer secured |
+| Data extraction | +4 weeks | Screening consensus |
+| Analysis + draft | +6 weeks | Data lock |
+| Supervisor review | +8 weeks | Draft complete |
+| Submission | +10 weeks | Supervisor approval |
+
+## Data Sources Used
+- PubMed E-utilities: ✅ (esearch count + efetch metadata)
+- Consensus MCP: ✅/❌
+- Scholar Gateway: ✅/❌
+- bioRxiv/medRxiv: ✅/❌
+- PROSPERO: ✅
+````
+
+## Solo-Mode Adaptations (no supervisor)
+
+When a user runs `/ma-scout` in topic-first mode without a supervisor assignment,
+substitute:
+- `## Overview` — drop the `Supervisor`, `Supervisor's area`, and timeline `Supervisor proposal` /
+  `Supervisor review` rows.
+- Replace `Supervisor approval` in the Timeline `Prerequisite` column with `self-approval` or
+  the lead author's own sign-off step.
+- `## Professor's Authority` — rename to `## Author Authority` and describe the
+  user's own prior publications that justify leading this MA.

package/skills/ma-scout/references/project_readme_template_ko.md

@@ -0,0 +1,82 @@
+# MA-Scout Project README Template — Korean variant (PROSPERO-ready)
+
+> Opt-in Korean variant. The default is the English `project_readme_template.md`. Use this when the
+> README is supervisor / professor-facing for a Korean PI review. English section headings and
+> PICO/PIRD structure are kept (so the README matches downstream `/meta-analysis` and `/search-lit`
+> expectations); Korean prompts are used for the PI-facing fields.
+
+Copy the block below into `{topic_folder}/README.md` and fill in the curly-brace fields.
+
+````markdown
+# {Topic Title} — {MA Type} Meta-analysis ({교수님 성함})
+
+## Overview
+- Supervisor: {교수님 성함} ({소속 이력})
+- 교수님 영역: Pillar {N} — {영역명}
+- Status: 기획 단계
+- Priority: {순위}순위
+- Created: {YYYY-MM-DD}
+
+## Research Question
+### PICO/PIRD
+- **P**opulation: {구체적 환자군, 질환, 세팅}
+- **I**ndex test / Intervention: {검사법 또는 중재}
+- **C**omparator / Reference standard: {비교군 또는 참조표준}
+- **O**utcome: {DTA: Se/Sp/AUC, Prognostic: HR/OR, 부작용 등}
+
+### One-line RQ
+{완성형 연구 질문 1문장}
+
+## Key Gap
+- 기존 MA: {N}편 ({상세, 가장 최근 연도, 범위 한계})
+- Consensus/Scholar Gateway 추가 확인: {결과 요약}
+- medRxiv/bioRxiv preprint MA: {있음/없음}
+- PROSPERO 등록 프로토콜: {있음 (CRD#) / 없음}
+- {구체적 gap 설명 — 왜 새 MA가 필요한지}
+
+## Professor's Authority
+- {분야 관련 논문 수, 대표 논문 1-2편, 독보적 기여}
+- {왜 이 교수님이 이 주제에 적합한지}
+
+## Preliminary Search ({날짜})
+### Search Strategy (PubMed)
+```
+{실제 사용한 PubMed 검색식 — E-utilities esearch query 그대로}
+```
+- Total hits: {N}편 (raw)
+- DTA/outcome extractable (estimated): {N}편 (×0.15-0.30 discount)
+- 기존 MA: {N}편 (narrow) / {N}편 (broad SR 포함)
+- Consensus 검색 결과: {N}편 추가 발견 여부
+- bioRxiv/medRxiv: {N}편 preprint
+
+### Embase Search Strategy (Draft)
+```
+{Embase용 검색식 초안 — Emtree 용어 포함}
+```
+(실행 전 초안 — Embase 접속 시 검증 필요)
+
+## Target Journal
+| 순위 | 저널명 | IF | MA 게재 비율 | Turnaround | 비고 |
+|------|--------|-----|-------------|-----------|------|
+| 1차 | {저널} | {IF} | {높음/중간} | {개월} | {근거} |
+| 2차 | {저널} | {IF} | {높음/중간} | {개월} | {근거} |
+
+## Timeline (역산)
+| 단계 | 예상 시점 | 선행 조건 |
+|------|----------|----------|
+| 교수님 제안 | {YYYY-MM} | {조건} |
+| PROSPERO 등록 | +1주 | 교수님 승인 |
+| 검색 완료 | +2주 | PROSPERO 등록 |
+| 스크리닝 완료 | +3주 | 2nd reviewer 확보 |
+| 데이터 추출 | +4주 | 스크리닝 합의 |
+| 분석 + 초안 | +6주 | 데이터 lock |
+| 교수님 리뷰 | +8주 | 초안 완성 |
+| 투고 | +10주 | 교수님 승인 |
+
+## Data Sources Used
+- PubMed E-utilities: ✅ (esearch count + efetch metadata)
+- Consensus MCP: ✅/❌
+- Scholar Gateway: ✅/❌
+- bioRxiv/medRxiv: ✅/❌
+- PROSPERO: ✅
+````

package/skills/ma-scout/skill.yml

@@ -0,0 +1,33 @@
+schema_version: 2
+name: ma-scout
+layer: D
+owner_domain: ma_topic_discovery
+
+when_to_use: "Discover and assess meta-analysis topics, either from a professor's publication profile or from a clinical question/trend."
+when_NOT_to_use: "Running the meta-analysis (use meta-analysis); single-author profiling (use author-strategy)."
+
+inputs:
+  - "professor profile or clinical question/trend"
+outputs:
+  - "ranked MA topic proposals with feasibility and gap evidence"
+side_effects:
+  - writes_report_artifacts
+  - network_access_literature
+downstream_consumers:
+  - meta-analysis
+  - search-lit
+forbidden_actions:
+  - claim_an_ma_gap_without_search_evidence
+  - fabricate_study_counts
+
+# v2.1 quality card
+purpose: "Find feasible meta-analysis gaps (professor-first or topic-first) backed by literature evidence."
+safety_boundaries:
+  - "Gap claims rest on real literature scans; candidate study counts are not fabricated."
+  - "Advisory report; does not perform screening or synthesis."
+known_limitations:
+  - "Feasibility is an estimate; the true includable pool is established only by meta-analysis screening."
+  - "No standalone demo; topic selection needs domain judgement."
+validation_commands:
+  - "confirm feasibility by running meta-analysis screening on the shortlist"
+evidence_surface: manual_workflow