Paper Reference Formatter

Common Challenges in Formatting Academic References

In academic writing and manuscript submission, formatting references is a tedious and error-prone task. Different journals, conferences, or institutions require various styles such as GB/T 7714, APA, MLA, or Chicago. Manual adjustment is not only time-consuming but can also introduce inconsistencies, affecting the professionalism of the paper. For instance, missing required fields or misclassifying reference types may lead to non-compliance with guidelines or even rejection.

Core Workflow of the Skill

This skill automates the reference formatting process through the following steps:

• Identify Output Style: First, confirm the desired format from the user. If not specified, the skill prompts a choice among GB/T 7714, APA, MLA, or Chicago.
• Parse References: Each reference is parsed into structured fields, including authors, title, year, volume/issue, pages, etc. Fields like DOI, URL, and access date are also extracted.
• Infer Reference Type: Based on clues, the type is conservatively inferred. For example:
• Journal article: from journal name, volume/issue/pages, DOI.
• Thesis/dissertation: from degree type, university, place, and keywords like "硕士/博士/学位论文" in Chinese or "dissertation/thesis" in English.
• Conference paper: from conference/proceedings name, location, date, editors, pages.
• Book: from publisher, place, edition, ISBN.
• Report/preprint/webpage: from report number, arXiv/bioRxiv/SSRN, URL, access date.
• Preserve User-Provided Facts: Do not invent missing bibliographic data. If a required field is missing, it is either omitted if the style allows or marked as [missing: field] if omission would be misleading.
• Normalization: Names, dates, titles, punctuation, and identifiers are normalized according to the chosen style. For Chinese references, Chinese names and titles are preserved unless the user requests translation.
• Output Formatted List: By default, a clean numbered list is generated. A short "Needs confirmation" note is added only for entries with unresolved missing or ambiguous fields.

Formatting Resources and Quality Checks

The skill provides auxiliary resources for accuracy:
- Style Rules File: Refer to references/style-rules.md for exact punctuation, field order, or type-specific templates.
- Formatting Script: When user data can be represented as JSON records, use scripts/format_reference.py for deterministic formatting. The script supports gbt7714, apa, mla, and chicago.

Quality checks include:
- Keeping each entry in the selected style only; no mixing of styles.
- Ensuring author order is unchanged from the input.
- For APA, MLA, and Chicago, normalizing DOI to https://doi.org/... when present.
- For GB/T 7714, including document type markers like [J], [D], [C], etc., where applicable.
- For APA, using sentence case for article/thesis titles when possible, while preserving proper nouns and original capitalization in Chinese titles.
- For MLA and Chicago, using headline-style title capitalization only when input supports it confidently; otherwise, preserving title capitalization.
- Reporting unresolved ambiguity briefly after the formatted list to avoid interrupting the flow.

Applicability and Caveats

This skill is suitable for handling common academic references but has limitations:
- It only supports GB/T 7714, APA, MLA, and Chicago styles. Other formats require manual adjustment.
- It relies on the quality of input data: if original reference information is incomplete, the output may include marked missing fields.
- For non-standard or highly specialized references, type inference may be inaccurate and require user verification.
- The script format_reference.py requires data in JSON format, ideal for batch processing; for interactive use, the skill guides step-by-step.

By automating formatting, this skill helps researchers reduce manual effort and focus on content creation, but users should still review the final output for accuracy.