
from chinese-reference-formatter-skill36
Formats Chinese and English literature titles into GB/T 7714-style Chinese bibliography entries and BibTeX, validating metadata against public sources before fo
This skill converts one or more Chinese or English reference titles into normalized Chinese-style bibliography entries (GB/T 7714) and BibTeX. It instructs the agent to search authoritative metadata sources (Crossref, DataCite, OpenAlex, Semantic Scholar, DBLP, PubMed, publisher pages, CNKI/Wanfang when available) and to prefer DOI/publisher metadata. If conflicts or insufficient evidence exist, the skill returns an ambiguity note rather than inventing metadata. The skill also normalizes metadata to a JSON schema and points to a formatting script for final rendering.
Use this skill when a user asks to: complete or format academic references from a title, generate a GB/T 7714-style reference list, create BibTeX entries from Chinese or English titles, or verify bibliographic metadata for inclusion in manuscripts or bibliographies. It's appropriate when accuracy and source verification are required.
scripts/format_reference.py hinted). (has_scripts=false in fetch output indicates none bundled.)This skill is implementation-neutral and works with agents that can run web lookups and invoke Python helpers (Copilot/Codex-style agents or other agents with internet access and Python runtime).
Chinese academic reference formatter following GB/T 7714 standard. Accepts normalized JSON metadata via stdin and outputs formatted Chinese bibliography entries plus BibTeX. The main formatter script requires piped JSON input (crashed without it in dry run, but this is expected). Both test suites passed cleanly. Code uses only Python stdlib — no external dependencies needed. Well-structured with clear type handling for articles, books, theses, patents, standards, and online resources.
Clean, well-written niche skill. No security concerns. format_reference.py exit code 1 in dry run is expected (no stdin). Tests are thorough and pass. CJK/western author handling is properly differentiated. BibTeX key generation uses author+year+title slug with MD5 fallback — functional if not elegant.