Survey Generation

作者 lingzhi2279e6c085d65e3無授權條款386 個星標收錄於 2026年10月8日更新於 2026年10月8日儲存庫7 個月前更新

Generate complete academic survey papers using multi-LLM parallel outline generation, RAG-based subsection writing, citation validation, and local coherence enhancement. Based on AutoSurvey pipeline. Use for writing comprehensive literature surveys.

AI 產生的概覽

產生完整學術綜述論文,包含多模型大綱、以 RAG 為基礎的撰寫、引用驗證與 BibTeX 輸出。

功能
此技能引導代理針對指定主題產出一篇完整的學術文獻綜述。它會從 Semantic Scholar 或 arXiv 蒐集 50 至 200 篇論文,透過平行大綱產生合併出整體大綱,搭配檢索到的文獻撰寫各小節並加入行內引用,使用 NLI 進行引用忠實度驗證,改善局部連貫性,並將引用轉換為 BibTeX。最終輸出一個綜述資料夾,內含 main.tex、references.bib、outline.json 以及各章節檔案。
適用情境
當需要針對某個研究主題從零撰寫一篇有引用支撐的完整文獻綜述時使用。它適合需要結構化大綱、可驗證參考文獻與 LaTeX 稿件的學術寫作任務。
執行需求
需要能執行 Python 腳本的代理執行環境,以及對 Semantic Scholar 或 arXiv 的網路存取;流程引用了位於 ~/.claude/skills/deep-research/scripts/search_semantic_scholar.py 的檢索腳本,以及 references/survey-prompts.md 中的提示範本。此技能本身不附帶腳本,只有指示文件與一個參考檔案。

Survey Generation

Generate complete academic survey papers with structured outline, RAG-based writing, and citation validation.

Input

  • $0 — Survey topic or research area

Scripts

Literature search

bash
python ~/.claude/skills/deep-research/scripts/search_semantic_scholar.py \  --query "relevant search query" --max-results 50

References

  • Survey prompts (outline, writing, citation, coherence): ~/.claude/skills/survey-generation/references/survey-prompts.md

Workflow (from AutoSurvey)

Step 1: Collect Papers

  1. Search Semantic Scholar / arXiv for papers on the topic
  2. Collect 50-200 relevant papers with titles and abstracts
  3. Filter by relevance and citation count

Step 2: Generate Outline (Multi-LLM Parallel)

  1. Generate N rough outlines independently (parallel)
  2. Merge outlines into a single comprehensive outline
  3. Expand each section into subsections
  4. Edit final outline to remove redundancies

Step 3: Write Subsections (RAG-Based)

For each subsection:

  1. Retrieve relevant papers for the subsection topic
  2. Generate content with inline citations [paper_title]
  3. Enforce minimum word count per subsection
  4. Only cite papers from the provided list

Step 4: Validate Citations

For each subsection:

  1. Check that cited paper titles are correct
  2. Verify citations support the claims made
  3. Remove or correct unsupported citations
  4. Use NLI (Natural Language Inference) for claim-source faithfulness

Step 5: Enhance Local Coherence

For each subsection:

  1. Read previous and following subsections
  2. Refine transitions and flow
  3. Preserve core content and citations
  4. Ensure smooth reading experience

Step 6: Convert Citations to BibTeX

  1. Replace [paper_title] with \cite{key}
  2. Generate BibTeX entries for all cited papers
  3. Validate all citation keys exist in .bib file

Output Structure

survey/├── main.tex          # Complete survey paper├── references.bib    # All citations├── outline.json      # Survey outline└── sections/         # Individual section files

Rules

  • Only cite papers from the collected paper list — never hallucinate citations
  • Each subsection must meet minimum word count
  • No duplicate subsections across sections
  • Citation validation is mandatory before final output
  • Local coherence enhancement must preserve all citations
  • The survey should be comprehensive and logically organized

Related Skills

  • Upstream: deep-research, literature-search, literature-review
  • See also: related-work-writing

來源與署名

來源:lingzhi227/agent-research-skills位於skills/survey-generation提交9e6c085

授權條款: 無授權條款

內容歸原作者所有。SourceWeft 從公開儲存庫中收錄這些內容。

檢舉或申請下架