zechang-xiong/chinese-reference-formatter-skill

Chinese Reference Formatter Skill

Organize “only titled” literature into verifiable Chinese paper references and BibTeX.

Example of effect · Installation · What does it solve? · Working principle · Development

Quick Installation:

It requires an agent capable of networked search, file access, and execution of Python; Formatter itself only relies on the Python 3.9+ standard library. See CHANGELOG.md for version updates.

---

In a word

The most annoying thing when manually organizing references is the formatting; When letting AI handle it, the biggest concern is unreliable information.

This skill solves both problems: first verify public metadata, then output the reference entry for the common Chinese paper and BibTeX. If no explicit reference is found, it will simply say it cannot be found; If there is a source conflict, it will provide ambiguous explanations.

Adjust the format less and do not include references.

Format and National Standard Version

Current 1.1.1 provides a common sequential coding reference format; for specific templates, see Format Rules. Not yet declared to fully comply with GB/T 7714—2015 or GB/T 7714—2025; The “standard references” in the output are result labels, not compliance certification.

GB/T 7714—2025 was implemented on July 1, 2026; the date is recorded in the official national standard records. When a university or journal specifies a version, its requirements prevail, and the Agent must adjust the output according to the relevant rules. Currently, it does not include the complete implementation of author-publication year, main citation management, or all document types for the 2025 edition.

---

Example of the effect

You enter

It outputs

The following content is generated directly by examples/metadata.json and is consistent with the script by test checks. GCN example references the ICLR meeting version, using the meeting page URL; arXiv DOI is a preprint record and is not used as a DOI for the meeting version.

bibtex @inproceedings{kipf-2017-semi-supervised-classification-graph-29814ac722ca, author = {Kipf, Thomas N and Welling, Max}, title = {{Semi-Supervised Classification with Graph Convolutional Networks}}, booktitle = {International Conference on Learning Representations}, year = {2017}, url = {https://openreview.net/forum?id=SJU4ayYgl} }

bibtex @inproceedings{lin-2017-focal-loss-dense-f15af3b59451,