search-arxiv-astrophysics

name: search-arxiv-astrophysics description: Search arXiv for astrophysics papers using user-provided and frequently changing research keywords, then return a ranked shortlist with concise relevance summaries. Use when a user asks to find recent or foundational astrophysics literature, monitor new papers for evolving interests, or explore broad research directions from keyword inputs.

Search Arxiv Astrophysics

Overview

Run repeatable arXiv searches for astrophysics topics from the user's current keyword list. Use the bundled script to fetch results, rank them by keyword match, and present a compact reading shortlist.

Workflow

1. Collect a fresh keyword set for this run.

• Ask for 5-15 terms or short phrases.
• Encourage coverage across object, process, method, and data context.
• Treat the keyword list as disposable and update it each session.

2. Build one or more focused queries.

• Use --match any for broad discovery and --match all for tighter filtering.
• Add --exclude terms to remove common false positives.
• Keep astrophysics scope using default categories unless the user requests otherwise.

3. Run the script.

python3 scripts/search_arxiv_astro.py \
  --keywords-file assets/keywords.txt \
  --match any \
  --days 365 \
  --max-results 40

Keyword file format supports:

• one term per line
• comma-separated terms on the same line
• comments with #

4. Present results for decision-making.

• Report top matches with title, date, category, and one-line relevance note.
• Highlight why each item matched the provided keywords.
• Separate "read now" from "background scan" papers.

Query Tactics For Broad Interests

When the user has broad or shifting interests, split search into 2-3 batches instead of one giant query:

• Batch A: science target keywords (objects or systems)
• Batch B: mechanism keywords (physical process)
• Batch C: method keywords (inference/modeling/instrument)

Run each batch and merge the top matches manually in the final summary. Read references/keyword-playbook.md for reusable keyword templates and category guidance.

Script Interface

• Script: scripts/search_arxiv_astro.py
• Required: at least one of --keyword (repeatable) or --keywords-file
• Common optional flags:
  • --keywords-file <path/to/keywords.txt>
  • --match any|all
  • --exclude <term> (repeatable)
  • --days <N> (recency filter after fetch)
  • --days-field published|updated|either (use updated to include revised papers)
  • --max-results <N>
  • --sort-by relevance|lastUpdatedDate|submittedDate
  • --output markdown|json
  • --insecure-tls (use only if local cert trust is broken)
  • --secure-tls (override env default and force verification)
  • --dry-run (print API URL without fetching)

Troubleshooting

• If HTTPS fails with certificate verification errors, retry with --insecure-tls.
• To include recently revised papers (older submission, new text), use:
  • --days <N> --days-field updated
• To make insecure TLS the default for your shell session:
  • export ARXIV_INSECURE_TLS=1
• To restore strict verification in one run:
  • add --secure-tls
• If the query is too broad, move generic terms to --exclude or split into batch runs.

Output Standard

For each recommended paper, include:

• Title and arXiv ID link
• Published date (ISO date)
• Primary category
• Matched keywords
• One-sentence relevance summary tied to the user's stated interests