SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsjournal-digest

journal-digest

Writing & research

View repository

Updated 10-15 17:30

About this skill

Top-Tier Economics and Finance Journal Literature Digest

Overview

This skill scrapes the latest published/online papers from top-tier journals in economics, finance, and accounting, filters literature related to corporate finance, generates a structured Chinese review weekly report, and proposes research topic suggestions.

Workflow

Step 1: Confirm Parameters

Confirm the following parameters with the user (use default values if the user does not specify):

Parameter Default Value Description
Time Range Last 7 days The time window for scraping papers; the user may specify e.g. "last two weeks", "March 2026"
Number of Selected Papers 15-25 papers The number of papers that ultimately enter the weekly report
Number of Topics 8-12 Number of research topic suggestions
Output Path F:\OneDrive\研究发展部\期刊速递\ Location where the output file is saved
Output File Name 公司金融周报_YYYY-MM-DD.md Named by generation date

Step 2: Scrape Papers

Run the scraping script to obtain the latest papers:

python "C:/Users/zhang/.claude/skills/journal-digest/scripts/fetch_papers.py" \
  --days <天数> \
  --output <临时JSON路径> \
  --history "F:/OneDrive/研究发展部/期刊速递/.paper_history.json" \
  --enrich

The script will:

  1. Scrape journals with RSS links from RSS sources
  2. For journals without RSS, obtain the latest papers through the Crossref API (including abstracts and author information)
  3. Use the --enrich parameter to supplement Crossref metadata for papers scraped via RSS but missing abstracts
  4. Compare with historical records to exclude papers that have already been scraped (deduplication)
  5. Output a candidate paper list (JSON format)

If the script fails to execute or some journals fail to be scraped, do not interrupt the process—continue with the successfully scraped papers. Note at the end of the weekly report which journals could not be successfully scraped this time.

If the script is unavailable or there are issues with the execution environment, you may fall back to manual mode: use the WebFetch tool to scrape journal RSS/webpages one by one and manually collect paper information.

Step 3: Filtering and Classification

Filter papers related to corporate finance from the candidate papers. Screening criteria:

CF-Core(directly within the field of corporate finance):

  • Corporate governance (boards, shareholder activism, voting rights, dual-class share structures)
  • Capital structure and financing decisions
  • Mergers and acquisitions (M&A motives, valuation, integration effects)
  • Dividend and share repurchase policies
  • Corporate investment decisions and capital allocation
  • Banks and financial intermediation (loans, deposits, regulation)
  • Corporate risk management
  • Executive compensation and incentives
  • IPOs and securities issuance
  • Private equity and venture capital
  • FinTech and corporate finance

CF-Cross(other fields that intersect with corporate finance):

  • Accounting information and corporate decision-making
  • ESG/climate finance and corporate behavior
  • Industrial organization and corporate strategy (antitrust, market structure)
  • Real Estate Finance
  • Labor Economics and Corporate Human Capital
  • Innovation and Intellectual Property
  • Supply Chain Finance
  • Digital Economy and Platform Firms

Methods/Data Worth Referencing：

  • The paper itself does not belong to corporate finance, but its methodology or data sources have important reference value for corporate finance research

Step 4: Generate Summaries

For each selected paper, generate a Chinese summary of about 150-250 characters, including:

  1. Paper Title(keep the original English title)
  2. Category Tags：[CF-核心]、[CF-交叉] or [方法/数据可借鉴]
  3. Source: journal name, volume/issue number or EarlyView/Ahead of Print designation, publication date
  4. Summary Body, organized according to the following structure:
    • Core research question (1 sentence)
    • Main findings/conclusions (2-3 sentences)
    • Overview of research methods and data (1-2 sentences)
    • Significance/implications for corporate finance research (1-2 sentences)
    • Author information (note if there are well-known scholars)

Summary writing requirements:

  • Use academic Chinese, concise and precise, avoid colloquial language
  • Highlight the paper's core contributions and innovations
  • Clearly identify the research methods (such as DID, IV, RDD, structural models, etc.)
  • If the paper involves causal identification, explain the identification strategy

Step 5: Generate Topic Suggestions

Based on the selected papers in this issue, propose 8-12 research topic suggestions. Each topic includes:

  1. Topic Title: a concise and clear research question
  2. Research Question: describe the core research question in 1-2 sentences; the context is not limited to China—it can be the United States, global, or cross-country comparison
  3. Data/Methods: suggested data sources and empirical methods
  4. Related Items: which papers in this issue it corresponds to

Principles for topic suggestions:

  • Topics should be feasible and data should be accessible
  • Prioritize topics with clear causal identification strategies
  • Encourage the use of quasi-natural experiments or institutional changes
  • The context can be China, the United States, Europe, or global/cross-country comparison
  • Topics should have a certain degree of innovation and not simply replicate the original paper

Step 6: Assemble Output

Assemble the final Markdown file according to the following template:

《公司金融周报》
2026年第XX周 | YYYY-MM-DD

---

## 一、本周精选论文（共XX条）

---

**1. [论文标题]**
[分类标签]
来源：期刊名, 卷期信息（时间）

[综述正文]

---

[... 更多论文 ...]

---

## 二、本周研究选题建议

---

**选题1：[选题标题]**
研究问题：[...]
数据/方法：[...]
关联条目：[...]

---

[... 更多选题 ...]

---

*本周报基于YYYY-MM-DD候选文献库（XX条）精选生成，覆盖[期刊列表]等顶级期刊。*

Step 7: Update History Records

Append the information of the papers selected this time to the history record file .paper_history.json, to ensure they will not be duplicated in the next retrieval.

History record format:

{
  "papers": [
    {
      "title": "论文标题",
      "doi": "DOI号（如有）",
      "journal": "期刊名",
      "date_fetched": "2026-04-15",
      "digest_file": "公司金融周报_2026-04-15.md"
    }
  ]
}

Journal List

The complete journal list and RSS configuration are available in references/journals.json。

Journals are divided by tier into:

  • Foreign A+: Top 5 Economics + Top 3 Finance
  • Foreign A: first-class journals
  • Foreign A-: excellent journals
  • Foreign B+: good journals

Retrieval priority: A+ > A > A- > B+. If there are too many candidate papers, prioritize keeping papers from higher-tier journals.

Notes

  • Paper titles retain the original English text and are not translated
  • Summaries are written in Chinese
  • If retrieval for a certain journal fails this time, note it at the end of the weekly report
  • Deduplication is based on a combined match of paper title + DOI, tolerating minor differences in the title (such as case and punctuation)
  • The weekly report number is calculated by the annual week number (ISO week number)