explore-recent-papers

name	explore-recent-papers
description	Downloads and deeply analyzes the full text of recent papers from an academic journal. Uses markitdown to convert PDFs to Markdown and reads the complete paper — not just abstracts and metadata.
license	MIT
compatibility	opencode
metadata	{"audience":"recent-papers-explorer","phase":"exploration","category":"papers"}

Goal

Download and analyze the full text of up to 20 recent publications from the target journal/conference. Read the complete papers — introduction, methodology, results, discussion — not just abstracts.

Search Strategy

Finding Recent Papers

1. Search arxiv for recent papers matching the journal name or conference proceedings
2. Search PubMed Central (open access) for biomedical journals
3. Search Semantic Scholar for papers with open-access PDFs
4. Search Google Scholar for the journal's recent publications
5. Check the journal's website for recent issue tables of contents
6. For each paper found, collect: title, authors, year, DOI, and PDF URL

Priority Order for Full-Text Access

For each paper, attempt to obtain the full text in this order:

1. arxiv preprint — https://arxiv.org/pdf/<arxiv-id>.pdf
2. bioRxiv / medRxiv preprint — use the PDF download link
3. PubMed Central — open-access versions of biomedical papers
4. Author's institutional page — many authors post preprints
5. Semantic Scholar — links to open-access PDFs
6. Journal open-access option — some papers are OA

If no full text is available after all attempts, fall back to abstract-only analysis for that paper.

Download Process

Create Paper Directory

mkdir -p "context/<journal-slug>/recent-papers"

Download PDFs

curl -L -o "context/<journal-slug>/recent-papers/<slugified-title>.pdf" "<pdf-url>"

Use a slugified version of the paper title (lowercase, hyphens, remove special characters) for filenames.

Convert PDFs to Markdown

markitdown "context/<journal-slug>/recent-papers/<slugified-title>.pdf" > "context/<journal-slug>/recent-papers/<slugified-title>.md"

After conversion, read the full Markdown text of each paper.

Verify Downloads

After downloading, verify each file:

ls -la "context/<journal-slug>/recent-papers/"

Check that files are not empty and have reasonable sizes (> 10KB for full papers).

Analysis Methodology

What to Extract from Full Papers

After reading the full text of each paper, analyze:

Topics & Themes

• What subjects are currently popular?
• What emerging trends exist?
• Which research gaps are being addressed?

Methodology Patterns

• Common research designs (experimental, observational, computational, etc.)
• Dominant data collection methods
• Popular analysis techniques and statistical approaches
• Typical sample sizes and study populations

Structural Conventions

• Typical paper sections and their organization
• Average paper length (in pages or word count)
• Abstract style (structured vs. unstructured, typical length)
• Figure and table usage patterns (how many, what types)

Citation Practices

• Citation style used by the journal (APA, Vancouver, etc.)
• Typical number of references
• Commonly cited frameworks, methods, or foundational works

Novelty Bar

• What counts as a sufficient contribution in this venue?
• Types of novelty valued (empirical, theoretical, methodological, practical)
• What is clearly insufficient for acceptance?

Sample Size

Aim for 20 papers with full text. If fewer full texts are available, supplement with abstract-only papers and clearly note the limitation.

Quality Checks

• Focus on RECENT publications (last 1-2 years)
• Verify papers are actually from the target journal
• Note how many papers were read in full vs. abstract-only
• Save all downloaded files for reproducibility

Output Location

Save to context/<journal-slug>/recent-papers-analysis.md (cache) AND <report-directory>/02-recent-papers-analysis.md (session artifact)

Keep all downloaded PDFs and converted Markdown files in context/<journal-slug>/recent-papers/ for future reference.