citation-management

Citation Management - Academic Citation Management and BibTeX Generation Tool

Skill Overview

Citation Management is a systematic citation management skill designed specifically for academic research. It supports searches on Google Scholar and PubMed, automatically extracts paper metadata, verifies citation accuracy, and generates BibTeX entries that comply with standard formats.

Use Cases

1. Paper Writing and Bibliography Construction

When writing academic papers, dissertations, or review articles, you may need to systematically find, organize, and verify references. This skill provides a complete workflow from literature retrieval to citation generation, ensuring the accuracy and formatting consistency of references while supporting the citation standards of high-impact journals and conferences such as Nature, Science, and NeurIPS.

2. Citation Format Conversion and Metadata Extraction

If you have a large number of DOIs, PMIDs, or arXiv IDs and need to convert them into complete BibTeX entries, this skill can automatically extract complete metadata—authors, title, journal, year, volume, issue, page numbers, and more—from authoritative sources such as CrossRef, PubMed, and arXiv, avoiding errors caused by manual entry.

3. Cleaning and Validating Existing Reference Libraries

If you already have a BibTeX file but it contains inconsistent formatting, duplicate entries, missing fields, or invalid DOIs, this skill provides formatting, deduplication, validation, and automatic repair functions to help you quickly clean and standardize your reference library.

Core Features

1. Academic Literature Search Across Multiple Databases

Supports automated searches on the two major databases, Google Scholar and PubMed. Google Scholar provides broad interdisciplinary coverage, while PubMed focuses on biomedical and life sciences literature (more than 35 million citations). Advanced features include year filtering, citation-count sorting, and MeSH term searching, helping you quickly find classic and influential papers in your field.

2. Automated Metadata Extraction

Given a paper identifier (DOI, PMID, arXiv ID, or URL), the system automatically extracts complete metadata from authoritative sources such as CrossRef, PubMed E-utilities, the arXiv API, and DataCite. Batch processing is supported, allowing multiple identifiers to be converted into structured BibTeX entries at once, greatly improving the efficiency of reference organization.

3. BibTeX Formatting and Validation

Provides standardized formatting for BibTeX files, sorting by year or keyword, duplicate-entry detection, DOI validity checks, required-field verification, and more. It supports citation-standard validation for different journals and conferences, including Nature, NeurIPS, and review articles, ensuring that references meet submission requirements.

Frequently Asked Questions

How does Citation Management help me search for academic papers?

This skill provides two search tools, search_google_scholar.py and search_pubmed.py, supporting keyword searches, year filtering, author filtering, citation-count sorting, and more. Google Scholar is suitable for broad interdisciplinary searches, while PubMed focuses on biomedical literature. Search results can be exported as JSON or directly converted to BibTeX format for subsequent organization and use.

How do I convert a DOI or PMID into BibTeX format?

Use the extract_metadata.py or doi_to_bibtex.py tool. For example, python scripts/doi_to_bibtex.py 10.1038/s41586-021-03819-2 automatically retrieves the DOI’s complete metadata from the CrossRef API and generates a BibTeX entry. The tool supports single-item conversion, batch conversion (reading multiple DOIs/PMIDs from a file), and extraction from arXiv IDs or URLs.

How can I verify that my references are accurate and complete?

Use the validate_citations.py tool to perform comprehensive validation of a BibTeX file, including whether DOIs are resolvable, whether required fields are complete, whether duplicate entries exist, and whether the formatting meets specifications. You can specify a target journal (such as --venue nature) to check whether the number of citations meets the relevant standard, or use the --manuscript parameter to check consistency between in-text citations and the BibTeX file, identifying uncited or missing references.

Which academic databases does this skill support?

It primarily supports Google Scholar (general interdisciplinary coverage) and PubMed (biomedical literature). Metadata extraction supports CrossRef (journal article DOIs), PubMed E-utilities (PMIDs), arXiv (preprints), and DataCite (datasets and software). Together, these sources cover the most commonly used literature resources in academic research.

How can I deduplicate and format a BibTeX file?

Use the format_bibtex.py tool. Adding the --deduplicate parameter automatically detects and removes duplicate entries. The tool also supports sorting by year or citation key, standardizing indentation and spacing, normalizing author-name formats, and fixing common errors such as using a single hyphen instead of -- for page ranges.

How many references are required when submitting to a high-impact journal?

This skill includes built-in citation-count validation standards for different journals and conferences. For example, Nature, Science, and Cell generally require 35–50 or more references; NeurIPS, ICML, and ICLR generally require 30–45 or more; and review articles generally require 40–65 or more. Use the --venue parameter to automatically check whether your reference count meets the standards of the target journal.

How is this skill different from other reference management tools?

Traditional tools such as EndNote and Zotero focus on graphical interfaces and manual management, whereas this skill focuses on automated scripts and workflow integration. It is better suited for users who need batch processing, custom scripts, or integration with LaTeX/Markdown writing workflows. Its command-line tools can be easily integrated into automated processes, making it particularly suitable for programmers and researchers who need to handle large numbers of references.

How can I find classic, highly cited papers in a particular field?

When searching Google Scholar, sort by citation count (--sort-by citations) to quickly find influential papers. This skill also provides citation-threshold guidelines: papers published 0–3 years ago with 20+ citations are worth following, papers published 3–7 years ago with 100+ citations are considered important contributions, and papers published more than 7 years ago with 500+ citations are regarded as classic works. Combining year filtering with citation-count sorting can help identify foundational literature in a field.