Hacktoberfest 2026：メンテナが10月に向けて印を付けた、オープンで初心者向けの issue。 Hacktoberfest の issue を見る

Expose DMN rule documentation and policy sources in the builder

オープン
  • 何か問題がありますか？
#563 コメント 0 件 リアクション 0 件 担当者 0 名 GitHub で見る

メンテナーはふだん 1 日以内に返信

まだ誰も着手していません。

評価

難易度
5/5
見積もり時間
1週間以上
初心者へのやさしさ
35/100
issue の種類
機能追加
明瞭さ
明確に書かれている
活発さ
活発
技術スタック
java, python
領域
backend-api-design, documentation, full-stack

調査の方向性

Start with .agents/skills/references/dmn-documentation.md, existing documented DMN models, and bin/library/load-library-metadata.py; trace how ModelRegistry extracts model descriptions and how builder-api stores and refreshes library metadata. Then follow the frontend library benefit/check views. Done means documentation and named sources reach the author details view with provenance, safe text rendering, older-metadata handling, and the specified round-trip, refresh, and frontend tests.

索引モデルが issue の本文から書いたものです。

説明

Library benefits and checks now carry their rule documentation in DMN descriptions, KIE documentation links (kie:attachment), shared type/BKM descriptions, and selective diagram text annotations. The builder currently receives model summaries and input/composition metadata, but authors cannot inspect the detailed rules, policy sources, review/effective dates, or known limitations without opening the DMN files.

Expose this documentation in the builder using the DMN as the source of truth. For example, an author selecting PECO CAP should be able to read why re-entry requires provider approval and open the cited handbook/order. Senior Citizen Tax Freeze should visibly disclose that its current model omits the program's income check.

Scope

  • Extend library discovery/OpenAPI metadata to represent model and decision documentation, named source links, and relevant annotations. Include descriptions for parameters and shared input types/fields, with references to their owning model/element rather than copying a separate policy document into every benefit.
  • Carry this metadata through bin/library/load-library-metadata.py, builder-api library metadata storage/refresh, and frontend library benefit/check views.
  • Provide a readable documentation/details view when choosing or reviewing library benefits and checks. Show rule interpretation, boundaries and unknown-input behavior, policy review/effective dates when present, named clickable sources, and implementation limitations.
  • Preserve provenance (model path, element ID, library version) so imported benefits/checks can identify their documentation source. Define refresh behavior and handle older metadata that lacks documentation.
  • Render descriptions as text and support the named documentation URLs without treating arbitrary DMN text as executable HTML.

Implementation notes

  • DMN 1.6 uses dmn:description; KIE Tools 10.2 links use dmn:extensionElements/kie:attachment with name and url. Notes use standard dmn:textAnnotation and optional associations to the relevant node.
  • ModelRegistry currently extracts only the model description. Inspect the original DMN XML for KIE extensions: the default Java engine marshaller skips extension children without registered converters, so extracting only the compiled runtime model can lose source links. The KIE Tools 10.2 editor marshaller preserves them.
  • Keep this read-only author documentation separate from executable decisions, applicant answers and eligibility results. Do not introduce an evaluated documentation decision.
  • The repository convention is .agents/skills/references/dmn-documentation.md; use existing documented models as fixtures, including shared fields and associated annotations.

Acceptance criteria

  • Benefit/check details show their DMN documentation and named policy sources without manually maintained duplicate content.
  • Shared input/parameter explanations are reachable from the relevant author workflow, with provenance intact.
  • Important caveats remain discoverable, including PECO CAP provider review and the current Senior Citizen Tax Freeze income gap.
  • Metadata round-trip/refresh tests cover root and decision links, shared type descriptions, associated annotations and models with no documentation.
  • Frontend tests cover readable details, source links, missing metadata and refresh behavior.

Related: #341 (library-check documentation), #549 (utility benefit models). This is distinct from #553/#554: those introduce applicant-facing application/help next steps and copy them into screeners; this issue exposes model/rule documentation for screener authors.

主要言語
Java
スター
16
フォーク
5
平均マージ
14時間 34分
マージ済み PR（30日）
23

環境構築

Codespaces で開く

このプロジェクトの開発コンテナを、あなたの GitHub アカウントでブラウザ上に起動します。

  • Dockerfile・Docker Compose ファイルなし
  • プルリクエストのテンプレートあり
  • コントリビューションガイドなし

はじめの一歩

  1. issue を最後まで読み、次にプロジェクトのコントリビューションガイドを読みます。
  2. 着手することを issue にコメントします — 二人が同じ作業をするのを防げます。
  3. リポジトリをフォークし、ブランチを切って変更します。
  4. issue 番号を参照したプルリクエストを送ります。

CodeForPhilly/benefit-decision-toolkit のほかの issue

CodeForPhilly/benefit-decision-toolkit の issue をすべて見る

似ている issue

Java の issue をもっと見る

新しい issue をメールで受け取る

初心者向けの GitHub issue を短くまとめたダイジェスト。