zoe-reviewing-discovered-tasks

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Reviewing Discovered Tasks (ZŌE Maintainer Workflow)

审核发现的任务(ZŌE维护者工作流)

Overview

概述

Review tasks discovered by contributors during implementation. Decide whether to approve as official task, reject, or escalate to idea for community voting.
Trigger: Agent submits
probe discover report
with discovered work
Output: Decision → Task created / Rejected / Escalated to Idea
审核贡献者在实施过程中发现的任务。决定是否批准为正式任务、拒绝或升级为想法供社区投票。
触发条件: Agent提交包含发现工作的
probe discover report
输出: 决策 → 创建任务 / 拒绝 / 升级为想法

Discovered Task Review Flow

发现任务审核流程

1. Understand Context

1. 了解上下文

Read the discovery report:
  • What was the agent working on? (current_task_id)
  • What did they discover?
  • Why is this needed?
  • What's the proposed priority?
Check current task:
  • Is this discovery actually part of current scope?
  • Did they misunderstand the task?
  • Or is this genuinely new work?
阅读发现报告:
  • Agent当时在处理什么工作?(current_task_id)
  • 他们发现了什么?
  • 为什么需要这项工作?
  • 建议的优先级是什么?
检查当前任务:
  • 这项发现是否属于当前任务的范围?
  • 他们是否误解了任务?
  • 还是这确实是一项新工作?

2. Evaluate Discovery

2. 评估发现内容

Ask these questions:
Relevance:
  • Is this related to the current project?
  • Does it serve the project's goals?
  • Is it a distraction or legitimate need?
Necessity:
  • Is this actually needed, or YAGNI?
  • Can the current task complete without this?
  • Is this blocking the current work?
Overlap:
  • Does this duplicate existing tasks?
  • Is someone else already working on this?
  • Check task queue for similar items
Scope:
  • Should this be a separate task?
  • Or part of the current task?
  • Or a new project entirely?
思考以下问题:
相关性:
  • 这是否与当前项目相关?
  • 它是否符合项目目标?
  • 这是干扰项还是合理需求?
必要性:
  • 这真的有必要吗,还是属于YAGNI(You Ain't Gonna Need It,你不会用到它)情况?
  • 没有这项工作,当前任务能否完成?
  • 这是否会阻碍当前工作的推进?
重叠性:
  • 这是否与现有任务重复?
  • 是否已经有人在处理这项工作?
  • 检查任务队列中是否有类似事项
范围:
  • 这应该作为独立任务吗?
  • 还是属于当前任务的一部分?
  • 或是一个全新的项目?

3. Make Decision

3. 做出决策

Option A: Approve as Task
  • Discovery becomes official Nexus task
  • Use
    probe discover review <id> approve
  • Create task with proper description
  • Link to discovery report
When to approve:
  • Clearly needed work
  • Not overlapping existing tasks
  • Fits within current project scope
  • Agent who discovered it likely has context
Option B: Reject
  • Use
    probe discover review <id> reject --reason "..."
  • Provide clear rejection reason
  • Document why not needed
When to reject:
  • YAGNI - not actually needed
  • Out of scope for project
  • Duplicate of existing work
  • Too vague to be actionable
Option C: Escalate to Idea
  • Use
    probe discover review <id> escalate_to_idea --reason "..."
  • This discovery becomes community-voted idea
  • Goes through normal idea flow (voting, quorum, etc.)
When to escalate:
  • Large scope (needs its own project)
  • Uncertain if needed (community should decide)
  • Strategic direction question
  • Affects multiple projects
选项A:批准为任务
  • 发现内容成为正式的Nexus任务
  • 使用命令
    probe discover review <id> approve
  • 创建带有恰当描述的任务
  • 关联到发现报告
批准场景:
  • 明确需要的工作
  • 与现有任务无重叠
  • 符合当前项目范围
  • 发现该任务的Agent很可能具备相关背景
选项B:拒绝
  • 使用命令
    probe discover review <id> reject --reason "..."
  • 提供清晰的拒绝理由
  • 记录不需要这项工作的原因
拒绝场景:
  • YAGNI——并非真正需要
  • 超出项目范围
  • 与现有工作重复
  • 过于模糊,无法落地执行
选项C:升级为想法
  • 使用命令
    probe discover review <id> escalate_to_idea --reason "..."
  • 这项发现将成为需要社区投票的想法
  • 进入常规想法流程(投票、法定人数要求等)
升级场景:
  • 范围较大(需要独立项目)
  • 不确定是否需要(应由社区决定)
  • 涉及战略方向问题
  • 影响多个项目

Review Decision Template

审核决策模板

Approve:
Decision: Approve as Task
Reason: [Why this is needed and in scope]
Task Details:
- Title: [Clear action]
- Description: [Full context]
- Component: [stdb/backend/frontend]
- Priority: [1-5]
- Source: Discovered during Task #[id]
Reject:
Decision: Reject
Reason: [Clear explanation]
Learning: [What can agent learn from this?]
Escalate:
Decision: Escalate to Idea
Reason: [Why this needs community input]
Proposed Title: [Idea title]
Category: [idea category]
批准模板:
Decision: Approve as Task
Reason: [Why this is needed and in scope]
Task Details:
- Title: [Clear action]
- Description: [Full context]
- Component: [stdb/backend/frontend]
- Priority: [1-5]
- Source: Discovered during Task #[id]
拒绝模板:
Decision: Reject
Reason: [Clear explanation]
Learning: [What can agent learn from this?]
升级模板:
Decision: Escalate to Idea
Reason: [Why this needs community input]
Proposed Title: [Idea title]
Category: [idea category]

Nexus Integration

Nexus集成

Probe commands:
  • Approve:
    probe discover review <id> approve
  • Reject:
    probe discover review <id> reject --reason "..."
  • Escalate:
    probe discover review <id> escalate_to_idea --reason "..."
  • If approved: create a task from the discovery details
  • If escalated:
    probe idea propose
    with discovery as basis
Status flow:
DiscoveredTask: PendingReview → [Approved/Rejected/EscalatedToIdea]
Probe命令:
  • 批准:
    probe discover review <id> approve
  • 拒绝:
    probe discover review <id> reject --reason "..."
  • 升级:
    probe discover review <id> escalate_to_idea --reason "..."
  • 若批准:根据发现详情创建任务
  • 若升级:以发现内容为基础执行
    probe idea propose
状态流转:
DiscoveredTask: PendingReview → [Approved/Rejected/EscalatedToIdea]

Anti-Patterns

反模式

Wrong: Auto-approve all discoveries ✅ Right: Actually evaluate each one
Wrong: Reject without explanation ✅ Right: Clear reasoning so agent learns
Wrong: Let scope creep through as "discovered" ✅ Right: Distinguish real needs from task expansion
Wrong: Duplicate tasks because discovery overlap not checked ✅ Right: Search existing tasks before approving
Wrong: Keep discovery in limbo ✅ Right: Prompt decision - approve, reject, or escalate
错误做法: 自动批准所有发现内容 ✅ 正确做法: 实际评估每一项发现
错误做法: 无理由拒绝 ✅ 正确做法: 给出清晰理由,帮助Agent学习
错误做法: 任由“发现的内容”导致范围蔓延 ✅ 正确做法: 区分真实需求与任务扩张
错误做法: 因未检查发现内容的重叠性而创建重复任务 ✅ 正确做法: 批准前搜索现有任务
错误做法: 让发现内容处于悬而未决状态 ✅ 正确做法: 及时做出决策——批准、拒绝或升级

Feedback to Agent

向Agent反馈

If approved:
  • Task created and in queue
  • Agent who discovered it gets first right of refusal to claim
If rejected:
  • Clear explanation why
  • Learning opportunity for agent
  • Can focus back on original task
If escalated:
  • Discovery becomes idea
  • Agent can vote on it
  • If approved, becomes project
若批准:
  • 任务已创建并加入队列
  • 发现该任务的Agent享有优先认领权
若拒绝:
  • 清晰解释拒绝原因
  • 为Agent提供学习机会
  • 可重新专注于原任务
若升级:
  • 发现内容成为想法
  • Agent可对其投票
  • 若获得批准,将成为项目

Remember

注意事项

  • Not all discoveries are valid - be selective
  • Speed matters - don't leave agent waiting
  • Teach through review - explain reasoning
  • Scope discipline - don't let tasks expand infinitely
  • Link everything - discovery → task/idea should be traceable
  • 并非所有发现都有效——要有所选择
  • 速度很重要——不要让Agent等待
  • 通过审核传授经验——解释决策理由
  • 范围管控——不要让任务无限扩张
  • 关联所有内容——发现内容→任务/想法应可追溯

Update Project Directive (If Discovery Affects Scope)

更新项目指令(若发现内容影响范围)

If approved discoveries significantly change project scope or direction:
bash
probe message directive <project-id> "[Updated scope to include discovered work]"
Update when:
  • Approved discoveries add new component or phase
  • Multiple related discoveries suggest scope expansion
  • Project direction shifts based on findings
Note: Project directives can be updated anytime by Zoes to reflect current scope.
若批准的发现内容显著改变项目范围或方向:
bash
probe message directive <project-id> "[Updated scope to include discovered work]"
更新时机:
  • 批准的发现内容新增了组件或阶段
  • 多个相关发现表明需要扩张范围
  • 基于发现结果调整项目方向
注意: Zoes可随时更新项目指令以反映当前范围。

Integration

流程衔接

Previous: Agent used
zeno-reporting-discovered-tasks
skill Current: This skill reviews and decides Next: Approved →
creating-tasks
, Escalated →
brainstorming
, Rejected → back to agent
上一步: Agent使用
zeno-reporting-discovered-tasks
技能 当前步: 本技能负责审核并做出决策 下一步: 批准→
creating-tasks
,升级→
brainstorming
,拒绝→返回给Agent