citation-management

Run the doi_to_bibtex.py script with the DOI as an argument, and it queries the CrossRef API to return a formatted BibTeX entry. You can also pass a file of DOIs with --input for batch conversion or use --format json for structured output.

Use the search_pubmed.py script, which wraps the NCBI E-utilities API and supports MeSH terms, field tags, date ranges, and publication type filters. Results can be exported directly to JSON or BibTeX for bibliography building.

The extract_metadata.py script pulls from CrossRef for DOIs, PubMed E-utilities for PMIDs, and the arXiv API for preprint IDs. It automatically detects the identifier type, including parsing DOIs and PMIDs out of article URLs.

Yes, the validate_citations.py script detects duplicates by matching DOIs and flagging similar titles, while format_bibtex.py --deduplicate removes them automatically. It also catches preprint entries that should be updated to published versions.

API responses from CrossRef or PubMed sometimes return incomplete records, especially for online-first publications. The skill mandates a web-search enrichment phase using parallel_web.py to fill missing volume, pages, issue, or DOI fields before finalizing entries.

Yes, validate_citations.py verifies each DOI resolves via doi.org and cross-checks the returned CrossRef metadata against your BibTeX entry. Broken DOIs, metadata mismatches, and missing required fields are reported with severity levels in a JSON report.