本文へ移動
cccskills
無料GitHub で公開

academic-research

Create comprehensive academic research notes with deep literature coverage. Auto-detects language (EN prompt→EN output, TR prompt→TR output). Supports Obsidian markdown and PDF. Performs 8-15 iterative search cycles with 25-50+ sources for comprehensive coverage. Uses footnote citations and visual overviews.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md5.3 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Academic Research Skill

Deep literature research with 25-50+ sources through iterative search cycles.

Workflow

1. Auto-Detect & Clarify

Language: Auto-detected from user prompt (no need to ask)

Ask User:

Output Format:

  • "Obsidian markdown mı PDF mi?" / "Obsidian markdown or PDF?"

If Obsidian: "Vault'ta ilgili notları arayayım mı?" / "Search vault for related notes?" If yes: "Vault yolu?" / "Vault path?" → Default: /home/user/Documents/Obsidian Vault/

If PDF: Skip vault integration

Save Location: "Nereye kaydedeyim?" / "Where to save?"

Topic Analysis & Mode Suggestion:

  • Narrow/Specific → Focused: 12-18 sources, 5-7 search cycles
  • Broad/Complex → Comprehensive: 25-35 sources, 8-12 cycles
  • Very Broad → Deep: 40-50+ sources, 12-15 cycles
  • Comparative → Synthesis: 15-25 sources, 6-9 cycles

State suggestion, proceed unless user objects.

Scope: "Odak veya hariç tutulacak alan?" / "Focus or exclusions?"

2. Iterative Research Process

Search Cycles Based on Mode:

  • Focused: 5-7 iterative cycles
  • Comprehensive: 8-12 iterative cycles
  • Deep: 12-15+ iterative cycles
  • Synthesis: 6-9 iterative cycles

Each Cycle:

  1. Execute search with specific query
  2. Evaluate results (quality, relevance, gaps)
  3. Identify what's missing (perspectives, time periods, methodologies)
  4. Formulate next targeted query
  5. Cross-validate findings
  6. Document effective vs. ineffective strategies
  7. Repeat until comprehensive coverage

Research Behaviors:

  • Cycle 1-2: Broad searches → Key concepts, seminal works, major authors
  • Cycle 3-5: Targeted searches → High-impact papers, meta-analyses, reviews
  • Cycle 6-8: Gap filling → Missing perspectives, alternative views, contradictions
  • Cycle 9-11: Cross-validation → Verify contested claims, find consensus
  • Cycle 12+: Final sweep → Recent work, emerging trends, practical applications

Source Diversity Targets:

  • Types: 60%+ peer-reviewed articles, 10-15% meta-analyses/reviews, 5-10% books, rest varied
  • Temporal: Balance seminal works + recent (50%+ within 5 years for active fields)
  • Geographic: Multiple regions/countries when possible
  • Perspectives: Supporting + Critical + Alternative views

Quality Priority: Peer-reviewed > Meta-analyses > Conference > Preprints

Search in English for best global sources (translate findings if Turkish output in a formal, academic language)

3. Generate Output

File Naming:

  • Descriptive, Title Case
  • EN: Cognitive Load Theory.md
  • TR: Bilişsel Yük Kuramı.md

For Obsidian:

  • Use appropriate template (en/tr)
  • Footnote citations: text[^1] → [^1]: Full citation 🟢
  • Visual overview (mindmap)
  • Wikilinks to related notes (if vault provided)
  • MOC integration
  • Tag suggestions (no #)
  • Dataview metadata
  • Support with basic TikZ whenever logical

For PDF:

  • PDF-optimized template
  • Footnote citations
  • Visual overview
  • Professional formatting
  • No wikilinks

Quality Standards by Mode:

ModeSourcesCycles🟢 High-ImpactTablesDiagrams
Focused12-185-750%+1+1+
Comprehensive25-358-1260%+2+2+
Deep40-50+12-15+65%+3+3+
Synthesis15-256-955%+3+1+

4. Present Output

  1. Brief summary (2-3 sentences of key findings)
  2. Share file via present_files
  3. Mention: source count, quality distribution, confidence score
  4. Highlight: notable gaps, debates, or surprising findings
  5. List: suggested next research steps
  6. If Obsidian: mention related notes linked

5. Continue Research

If user says "continue", "expand", or "more sources":

  1. Load existing note
  2. Review research log → identify gaps
  3. Ask which gap to address
  4. Execute 3-5 additional cycles
  5. Integrate findings (avoid duplication)
  6. Update metadata (source_count, last_updated)
  7. Present updated note with summary

Research Depth Indicators

Track and report in note:

  • Total search cycles executed
  • Total sources found vs. included
  • Quality distribution (🟢🟡🔴 counts)
  • Temporal coverage (year range, % recent)
  • Geographic coverage (countries/regions)
  • Methodological diversity (RCT, meta, qualitative, etc.)
  • Perspective balance (supporting/critical/alternative)

Key Features

Deep Coverage:

  • 8-15 iterative search cycles
  • 25-50+ sources (comprehensive/deep modes)
  • Multiple perspectives required
  • Gap identification and filling
  • Cross-validation of claims

Smart Integration:

  • Auto language detection
  • Format choice (Obsidian/PDF)
  • Optional vault search
  • Footnote citations
  • Visual overviews

Quality Tracking:

  • Source diversity metrics
  • Confidence scoring (⭐1-5)
  • Research cycle documentation
  • Methodology notes

References

Templates: references/template-{en|tr}.md Citation: references/apa7-quick-ref.md Diagrams: references/mermaid-patterns.md Search: references/search-strategies.md

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Master backend architectures, REST API design, microservices, and server-side development across multiple languages. Use this as a central index to access specialized sub-skills.

日本語の概要は準備中です。原文の説明を表示しています。

MikeCheng1208/BattleTree22026年7月22日 更新

在每次回复中使用固定前缀并列出本次实际参考的指令/文件,便于校验是否遵循上下文与规则。

日本語の概要は準備中です。原文の説明を表示しています。

MikeCheng1208/BattleTree22026年7月22日 更新

Automate 21risk tasks via Rube MCP (Composio). Always search tools first for current schemas.

日本語の概要は準備中です。原文の説明を表示しています。

MikeCheng1208/BattleTree22026年7月22日 更新

Automate 2chat tasks via Rube MCP (Composio). Always search tools first for current schemas.

日本語の概要は準備中です。原文の説明を表示しています。

MikeCheng1208/BattleTree22026年7月22日 更新

Arquitecto de Soluciones Principal y Consultor Tecnológico de Andru.ia. Diagnostica y traza la hoja de ruta óptima para proyectos de IA en español.

日本語の概要は準備中です。原文の説明を表示しています。

MikeCheng1208/BattleTree22026年7月22日 更新

[00] META. Orchestrate skills into dynamic chains for complex tasks. Analyzes the task, discovers available skills, builds an optimal chain, explains why each skill is needed, and executes step-by-step with user confirmation. Use for any complex task requiring multiple thinking/research/analysis steps. Triggers on "plan this", "how to approach", "what's the strategy", "build a plan", or any multi-step problem requiring skill orchestration.

日本語の概要は準備中です。原文の説明を表示しています。

MikeCheng1208/BattleTree22026年7月22日 更新

MikeCheng1208 のスキルをすべて見る

このスキルの問題を報告する