Ccc Literature Scout

CCC Literature Scout & Introduction Synthesizer (V2.0 - Massive Scale)

A specialized academic skill for large-scale literature retrieval (up to 1000 papers per keyword), per-keyword .bib generation with full abstracts, multi-bib deduplication and synthesis, and publication-grade 5-stage Introduction drafting (Nature / Elsevier / IEEE style).

🌟 Two-Stage Systematic Architecture

[User Keywords: KW1, KW2, KW3...]
               │
               ▼
┌────────────────────────────────────────────────────────────────────────┐
│ STAGE 1: Massive-Scale Multi-Source Literature Retrieval               │
│ - OpenAlex (250M+ Works, Inverted-Index Abstract Reconstruction)       │
│ - Web of Science Clarivate API (Starter / Expanded API)                │
│ - Campus WoS Batch Exports (.ciw, .txt, .bib, .ris folder)             │
│ - arXiv & Crossref API Enrichment                                      │
│                                                                        │
│ OUTPUT per Keyword:                                                    │
│ ├── bib_by_keyword/01_<keyword1>.bib (up to 1000 entries with abstracts)│
│ ├── bib_by_keyword/01_<keyword1>_matrix.csv                            │
│ ├── bib_by_keyword/02_<keyword2>.bib (up to 1000 entries with abstracts)│
│ └── bib_by_keyword/02_<keyword2>_matrix.csv                            │
└────────────────────────────────────────────────────────────────────────┘
               │
               ▼
┌────────────────────────────────────────────────────────────────────────┐
│ STAGE 2: Deep Cross-Corpus Synthesis & Introduction Generation         │
│ - Ingest all keyword bib files into a unified corpus                   │
│ - Multi-level deduplication (DOI, Title cosine, arXiv mapping)         │
│ - Timeline evolution & Landmark citation ranking                       │
│ - Thematic clustering into methodological streams                      │
│                                                                        │
│ OUTPUT Master Artifacts:                                               │
│ ├── master_unified_references.bib (Clean, deduplicated master library) │
│ ├── master_literature_matrix.csv (Excel-ready analysis spreadsheet)   │
│ ├── master_literature_synthesis.md (Trend report & landmark ranking)   │
│ ├── introduction_draft.tex (Publication-grade LaTeX with \cite{...})   │
│ └── introduction_draft.md (Markdown draft for fast reading)            │
└────────────────────────────────────────────────────────────────────────┘

🛠️ CLI Quick Start & Commands

The tool is located at: skills/ccc-literature-scout/ccc_literature_tool.py

1. Full End-to-End Pipeline (Stage 1 + Stage 2)

Downloads up to 1000 papers with abstracts for each keyword, exports separate keyword bib files, and generates the unified introduction:

# English journal mode (Elsevier / Nature / IEEE)
python skills/ccc-literature-scout/ccc_literature_tool.py pipeline \
  --keywords "implicit neural representations microstructure" "porous media transport permeability" "multigrid defect correction" \
  --topic "Implicit Neural Microstructure Representation for Multiscale Transport" \
  --method-name "the proposed hierarchical implicit framework" \
  --limit 1000 \
  --output-dir ./literature_project \
  --lang en

# Chinese manuscript mode (中文期刊 / 博士论文引言)
python skills/ccc-literature-scout/ccc_literature_tool.py pipeline \
  --keywords "水泥基材料 微观结构 传输性能" "隐式神经表征 偏微分方程求解" \
  --topic "基于分层隐式神经表征的水泥基材料多尺度传输建模" \
  --method-name "分层隐式连续表征与缺陷修正物理求解器" \
  --limit 1000 \
  --output-dir ./literature_project_zh \
  --lang zh

2. Stage 1 ONLY: Batch Retrieval (1 Bib per Keyword)

python skills/ccc-literature-scout/ccc_literature_tool.py stage1-fetch \
  --keywords "deep learning fluid dynamics" "physics informed neural networks" \
  --limit 1000 \
  --output-dir ./stage1_data

3. Stage 2 ONLY: Synthesize Existing .bib Directory into Introduction

python skills/ccc-literature-scout/ccc_literature_tool.py stage2-synthesize \
  --bib-dir ./stage1_data/bib_by_keyword \
  --topic "Physics-informed neural networks for fluid dynamics" \
  --method-name "our conservative PINN architecture" \
  --output-dir ./stage2_output \
  --lang en

4. Direct Web of Science Clarivate API or Campus Export Ingestion

• If you have a Web of Science API Key:

  python ccc_literature_tool.py pipeline --keywords "..." --topic "..." --wos-api-key "YOUR_KEY"
  

• If you have a folder of exported .ciw, .txt, .bib files from campus WoS:

  python ccc_literature_tool.py pipeline --keywords "..." --topic "..." --wos-dir ./campus_wos_downloads
  

📂 Output Artifacts Directory Structure

literature_project/
├── bib_by_keyword/
│   ├── 01_implicit_neural_representations_microstructure.bib      (1000 entries with abstracts)
│   ├── 01_implicit_neural_representations_microstructure_matrix.csv
│   ├── 02_porous_media_transport_permeability.bib                 (1000 entries with abstracts)
│   ├── 02_porous_media_transport_permeability_matrix.csv
│   ├── 03_multigrid_defect_correction.bib                         (1000 entries with abstracts)
│   └── 03_multigrid_defect_correction_matrix.csv
├── master_unified_references.bib                                  (Deduplicated master bib)
├── master_literature_matrix.csv                                   (Excel-ready master matrix)
├── master_literature_corpus.json                                  (Structured JSON corpus)
├── master_literature_synthesis.md                                 (Timeline & landmark report)
├── introduction_draft.tex                                         (5-stage publication LaTeX)
└── introduction_draft.md                                          (Markdown preview)