Find the one paper you should read today from arXiv and Semantic Scholar
Read memory/MEMORY.md for context. Read the last 7 days of memory/logs/ to avoid recommending papers already covered.
Search for recent papers. Start with arXiv (no rate limits), then try Semantic Scholar as a supplement:
Primary — arXiv (always works, no rate limits):
# If ${var} is set, use it as the query. Otherwise use broad categories.
# arXiv categories: cs.AI, cs.CL (NLP), cs.LG (machine learning)
curl -s -L "https://export.arxiv.org/api/query?search_query=cat:cs.AI+OR+cat:cs.CL+OR+cat:cs.LG&sortBy=submittedDate&sortOrder=descending&max_results=15"
Secondary — Semantic Scholar (may 429, treat as optional):
curl -s "https://api.semanticscholar.org/graph/v1/paper/search?query=artificial+intelligence+large+language+models&year=2025-2026&limit=5&fields=title,authors,abstract,url,publicationDate,citationCount,openAccessPdf" \
-H "Accept: application/json"
If rate-limited (429), skip Semantic Scholar entirely — arXiv results are sufficient. Do not retry or wait.
If arXiv returned thin results or ${var} is a niche topic, also try WebSearch for "[topic] paper 2025 2026 site:arxiv.org" to catch papers the API missed.
From all results, pick the single best paper — the one most worth reading today. Criteria: novelty, relevance, practical implications. Skip anything already mentioned in recent logs.
Send via ./notify:
*Paper Pick — ${today}*
"Paper Title" — Authors
One sentence: why this paper is worth your time.
[Read](url) | [PDF](pdf_url)
If open-access PDF is available, include the PDF link. Otherwise just the paper link.
Log to memory/logs/${today}.md.
If nothing interesting found, log "PAPER_PICK_OK" and end.