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

[Feature]: Task board column for "agent done, waiting for the human's review"

クローズ
  • 何か問題がありますか？
#92 コメント 4 件 リアクション 0 件 担当者 0 名 GitHub で見る

まだ誰も着手していません。

評価

難易度
3/5
見積もり時間
1〜2日
初心者へのやさしさ
70/100
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
shell
領域
developer-experience, frontend

調査の方向性

Start with web/templates/tasks.html around line 498 and the task grouping logic in web/blueprints/tasks.py; trace how the board renders statuses and how partial completion and count_human_acs are computed. Add a derived Awaiting review grouping for started-work tasks with all Agent criteria ticked and at least one Human criterion unticked. Verify the board separates those tasks from In Progress without adding a status.

索引モデルが issue の本文から書いたものです。

説明

area:watchtower enhancement severity:low status:released triaged
Problem

The Watchtower task board cannot tell "the agent is working on it" from "it is waiting for my review".

The board has four columns, one per status: Captured, In Progress, Issues, Completed (web/templates/tasks.html, line 498). A task whose ### Agent criteria are all ticked and whose ### Human criterion is still open stays started-work, so it sits in In Progress next to the tasks the agent is still working on.

Observed: eleven cards in In Progress; for seven of them the agent's work was finished and only a Human criterion was left. Nothing on the board distinguished the seven from the four. The operator asked: "can we add a column to tell in progress for you from in review for me?"

Proposed Solution

Derive a fifth column, for example "Awaiting review", without adding a status: a task is shown there when its status is started-work, all ### Agent criteria are ticked and at least one ### Human criterion is unticked. The framework already computes this state (the partial-complete case, and count_human_acs used in web/blueprints/tasks.py), so the board would only need to group on it.

A count of unticked Human criteria on each card would be a smaller step in the same direction.

Subsystem

Watchtower (Web UI)

Scope

Medium (few files, 1-4 hours)

Alternatives Considered
  • Workaround in use: set owner: human on every task whose agent work is finished, and use the owner label and the owner=human filter. It works, but it overloads owner (it now means "whose turn it is", not "who is accountable"), and nothing sets it automatically.
  • /approvals lists the pending Human criteria, but as a separate page, not as a position on the board.
Additional Context

Version: 1.8.6 (1ea10441f), vendored consumer. First seen on 1.8.3; web/templates/tasks.html and web/blueprints/tasks.py are byte-identical between the versions. Checked again on the served board of 1.8.6: four columns. The suggestion was not prototyped.

主要言語
Shell
スター
16
フォーク
6
平均マージ
6時間 24分
マージ済み PR（30日）
5

環境構築

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。

DimitriGeelen/agentic-engineering-framework のほかの issue

DimitriGeelen/agentic-engineering-framework の issue をすべて見る

似ている issue

Shell/Bash 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。