What problem does it solve?
Turning raw web search results into a usable bibliography is tedious and error-prone: search snippets contain untrusted text, duplicate DOIs, and missing years that corrupt BibTeX files. This Skill converts normalized multi-search-engine JSON into a clean, minimal BibTeX file automatically.
Core Features & Use Cases
- JSON to BibTeX Conversion: Reads a multi-search-engine JSON payload on stdin and emits one
@misc{} entry per result with deterministic keys (ref1, ref2, ...).
- Identifier Preservation: Extracts and normalizes real DOIs, arXiv IDs (modern and legacy), authors, corporate authors, and publication years when available; unknown years are omitted rather than invented.
- Deduplication and Sanitization: Collapses duplicate DOI records to a single entry and neutralizes control characters, embedded BibTeX fragments, and unbalanced braces from untrusted search text.
- Use Case: After running a literature search across multiple engines, pipe the normalized JSON results through this Skill to produce
paper/references.bib with provenance notes that downstream citation-integrity gates can audit.
Quick Start
Pipe the JSON output of your multi-engine paper search into this Skill to generate a sanitized references.bib file with DOI and arXiv metadata preserved.
