evidence-level-ranker
Evidence Level Ranker | 证据等级排序器
Task
Use this skill to rank papers by evidence strength, methodological quality, and citation priority within one explicit comparison framework.
This skill should identify what kind of evidence each paper provides, how much methodological trust it deserves, how much validation or corroboration it contains, and whether it should be treated as a high-priority anchor citation, context-setting citation, mechanistic support citation, or low-priority / caution citation.
This skill must not equate study design labels with true evidentiary value automatically. A meta-analysis is not automatically decisive, an RCT is not automatically well-conducted, a cohort is not automatically weak, and a mechanism study is not automatically non-informative. The skill must rank literature based on the combination of design family, execution quality, validation depth, bias control, and claim discipline.
This skill is especially useful when the user needs to:
- prioritize citations for a manuscript, review, protocol, or slide deck;
- compare reviews, observational studies, interventional studies, mechanism papers, omics studies, and validation studies in one framework;
- identify which papers are most suitable for supporting strong claims versus background framing;
- avoid treating flashy but fragile findings as top-tier evidence.