사용 방법
- 설치 명령을 복사하고 자신의 에이전트 환경에 설치하기 전에 소스 패키지를 확인하세요.
- 필요한 도구, 인증 정보, 지원 환경은 원본 문서에서 확인하세요.
- 입력과 원하는 산출물을 설명하고 출처와 결과를 대조하세요.
npx skills add https://github.com/yogsoth-ai/de-anthropocentric-research-engine --skill aggregate-ranking시작 전 확인
- 설치에는 모델 구독, API 한도, 파일 접근 권한이 포함되지 않습니다.
- 별점 수와 설치 수는 활동 지표이며 독립적인 품질·보안 검증이 아닙니다.
원본 Skill 문서
원본 저장소의 제목, 예시, 코드, 표, 링크, 이미지를 유지해 표시합니다.
aggregate-ranking
Purpose
Aggregate criterion or comparison results into an ordered recommendation under an explicit rule.
Input contract
yaml
required: [candidate_set, criterion_results, aggregation_rule]
optional: [tie_break_rule, missing_value_policy, uncertainty_annotations]
constraints: [criterion directions and scales must be declared; no silent imputation]Procedure
- Align candidate identifiers, criterion directions, units, and validity flags.
- Apply the supplied aggregation rule without changing weights or directions.
- Propagate missingness and uncertainty; apply the declared tie-break only after aggregation.
- Return ordered candidates with component contributions and recommendation status.
Output contract
yaml
produces: [ordered_recommendation, aggregate_scores, contribution_table, unresolved_comparisons]
delta_fields: [findings, evidence_updates, decisions, uncertainties]Quality gates
- Every ranked candidate has a traceable value for each required criterion or an explicit unresolved marker.
- Aggregation reproduces the supplied rule and preserves criterion direction.
- Ties and sensitivity to tie-breaks are reported.
Parameterization
Caller supplies candidate schema, criterion scales/directions, weights or aggregation formula, tie-break rule, and missing/uncertainty policy.
Failure and counterexamples
Reject mixed units without normalization; reject a recommendation when a hard criterion is unresolved.
Provenance map
- resolved: priority-synthesis
- resolved: scoring-synthesis
Preserved source criteria ledger
| source | criterion |
|---|---|
| priority-synthesis | All scoring dimensions are present for every gap. |
| priority-synthesis | Weight vector sums to 1.0 within +/-0.001. |
| priority-synthesis | Priority list is sorted descending; ties use feasibility sub-score. |
| priority-synthesis | Top N is N=min(3,total gaps) and includes attack-path suggestions. |
| scoring-synthesis | Final recommendation includes recommended alternative, confidence, key assumptions, and risk warnings. |
