paper-collector

Paper Collector

Collects academic papers about AI Agent Skills and saves them to papers/.

Workflow

1. Search for Relevant Papers

# Search arXiv for skill-related papers
curl -s "https://export.arxiv.org/api/query?search_query=ti:agent+skill+OR+ti:tool+agent+OR+ti:agent+framework&start=0&max_results=20&sortBy=submittedDate&sortOrder=descending" | \
python3 << 'PYEOF'
import sys
import re
from datetime import datetime

xml = sys.stdin.read()

# Extract entries
entries = re.findall(r'<entry>(.*?)</entry>', xml, re.DOTALL)

papers = []
for entry in entries:
    title = re.search(r'<title>(.*?)</title>', entry, re.DOTALL)
    summary = re.search(r'<summary>(.*?)</summary>', entry, re.DOTALL)
    link = re.search(r'<id>(.*?)</id>', entry)
    published = re.search(r'<published>(.*?)</published>', entry)
    
    if title and link:
        title_clean = ' '.join(title.group(1).split())
        summary_clean = ' '.join(summary.group(1).split())[:300] if summary else ""
        link_url = link.group(1)
        date = published.group(1)[:10] if published else ""
        
        papers.append({
            'title': title_clean,
            'summary': summary_clean,
            'url': link_url,
            'date': date
        })

print(f"Found {len(papers)} recent papers:\n")
for p in papers[:10]:
    print(f"[{p['date']}] {p['title']}")
    print(f"  {p['summary'][:150]}...")
    print()
PYEOF

2. Also Search Google Scholar / Semantic Scholar

# Search Semantic Scholar API (free, no auth needed)
curl -s "https://api.semanticscholar.org/graph/v1/paper/search?query=agent+skill+framework&limit=20&fields=title,abstract,year,citationCount,openAccessPdf,authors" | \
python3 << 'PYEOF'
import sys
import json

data = json.load(sys.stdin)
papers = data.get('data', [])

print(f"Found {len(papers)} papers on Semantic Scholar:\n")
for p in papers[:10]:
    title = p.get('title', '')
    year = p.get('year', '?')
    citations = p.get('citationCount', 0)
    abstract = p.get('abstract', '')[:200] if p.get('abstract') else ''
    pdf = p.get('openAccessPdf', {}).get('url', '') if p.get('openAccessPdf') else ''
    
    print(f"[{year}] {title} (cited: {citations})")
    if pdf:
        print(f"  PDF: {pdf}")
    print()
PYEOF