使い方
- インストールコマンドをコピーし、自分の環境へ導入する前にソースを確認します。
- 必要なツール、認証情報、対応環境を原文で確認します。
- 入力と成果物を指定し、出力を資料と照合します。
npx skills add https://github.com/yogsoth-ai/de-anthropocentric-research-engine --skill categorize-evidence開始前の確認
- 導入にはモデル契約、API枠、ファイルへのアクセス許可は含まれません。
- スター数や導入数は活動の指標であり、独立した品質・安全性試験ではありません。
リポジトリの原文
見出し、例、コード、表、リンク、参照画像を含む原文を表示しています。
categorize-evidence
Purpose
Cluster an evidence corpus by a declared thematic, methodological, chronological, mechanistic, or other schema.
Input contract
yaml
required: [evidence_records, category_schema]
optional: [coding_rules, multi_label_policy, seed_categories]
constraints: [category assignments require record evidence and allow explicit multi-label or unknown states]Procedure
- Define category semantics and assignment rules.
- Code each record using the declared evidence fields.
- Review boundary cases and preserve multi-label or unresolved assignments.
- Summarize category coverage and representative records.
If categorized sources now need claim-level structured records, consider extract-evidence-record as the next tactic.
Output contract
yaml
produces: [categorized_corpus, category_definitions, boundary_cases, coverage_summary]
delta_fields: [findings, evidence_updates, uncertainties, open_questions]Quality gates
- Categories are mutually interpretable even when not mutually exclusive.
- Every assignment is traceable to record fields.
Failure and counterexamples
Do not force records into categories whose definitions do not fit, and do not confuse frequency with evidential importance.
Provenance map
resolved: knowledge-acquisition-categorize-papersconcept: knowledge-structuring/source-categorization-patternsintermediate: Pass4/map-field-taxonomy
