Rank candidate source documents for text passages using lexical similarity.
npx skills add https://github.com/LegalQuants/lq-skills --skill text-provenance
Please help me install this Agent Skill. Skill: text-provenance Source: https://github.com/LegalQuants/lq-skills/tree/main/skills/text-provenance Command: npx skills add https://github.com/LegalQuants/lq-skills --skill text-provenance
Helps you identify the likely source document of a text passage, support clause attribution in legal documents, and flag potential plagiarism through fast lexical matching signals.
Use the text-provenance skill to compare a snippet against your corpus documents and return ranked provenance candidates with similarity-based confidence scores.
High-intent search queries and answers about installing and using this skill.
You can identify contract clause provenance by matching clauses against a playbook or standard clause origins using lightweight lexical similarity. The skill ranks candidate source passages with confidence scores to support derivation workflows.
Yes, you can generate RAG citation highlighting deterministically without runtime embeddings or API calls. The skill relies on surface similarity, n-gram overlap, and exact fingerprint matching to produce ranked provenance candidates.
Lexical similarity scoring for plagiarism screening works by calculating n-gram overlap and exact phrase matching to produce candidate matches. It operates deterministically to flag potential plagiarism for investigation without making definitive adjudications.
No, you do not need external API calls for text attribution matching. The skill operates deterministically using surface similarity and exact fingerprint matching, ensuring fast candidate provenance rankings without runtime dependencies.
A limitation of using surface similarity for contract analysis is that it produces candidate matches to investigate rather than definitive adjudications. It relies strictly on lexical overlap and fingerprint matching without deeper semantic understanding.
RecursiveIntellAdd ASCII borders and color styles to terminal text.
dailycafiMatch MS/MS spectra against reference libraries to identify compounds.
vishalsachdevIndexes documents and searches exact or fuzzy text matches with positional tracking.
qcmuuGenerate semantic embeddings from text using sentence-transformers models.
