Token导航 LogoToken导航

Agent Skill · 研究检索

pmo-retrospective(项目管理办公室回顾)

pmo-retrospective 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

pmo-retrospective 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 51,060。

总安装 594周安装 25GitHub Stars 180下载 208核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:pmo-retrospective(项目管理办公室回顾)
来源仓库:https://github.com/lerianstudio/ring
仓库路径:skills/pmo-retrospective
安装命令:
npx skills add https://github.com/lerianstudio/ring --skill pmo-retrospective
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/lerianstudio/ring --skill pmo-retrospective

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

pmo-retrospective 提供项目回顾框架与反合理化策略,帮助团队系统化总结经验、识别根本原因并跟进行动。适用于需要组织复盘会议、分析失败模式或建立学习机制的项目管理场景。通过结构化模板输出反思文档,支持无责备文化下的持续改进流程。安装前需确认权限范围与维护状态,注意可能触发文件读写与数据访问操作。

建议结合原始 README 核验具体用法及拦截标准,确保在合适时机启动回顾流程。

来源仓库内路径:skills/pmo-retrospective

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/lerianstudio/ring --skill pmo-retrospective;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

pmo-retrospective 怎么安装?

在终端执行 npx skills add https://github.com/lerianstudio/ring --skill pmo-retrospective。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

pmo-retrospective 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

pmo-retrospective 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

pmo-retrospective 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

pmo-retrospective 用的人多吗?

公开数据是累计安装 594 次,近一周安装 25 次,下载 208 次,GitHub 180 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

pmo-retrospective 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
pmo-retrospective

项目管理办公室回顾 · 本页

594180只读未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

PMO Retrospective Skill

Systematic capture and application of lessons learned at portfolio level.

Purpose

This skill provides a framework for:

  • Lessons learned capture
  • Pattern identification
  • Process improvement
  • Organizational learning
  • Knowledge transfer

Retrospective Types

Type 1: Project Closure Retrospective

Trigger: Project completion or termination Scope: Single project Participants: Project team, sponsor, key stakeholders


Type 2: Portfolio Period Review

Trigger: Quarterly or annual review Scope: All projects in period Participants: PMO, project managers, executives


Type 3: Thematic Retrospective

Trigger: Recurring pattern observed Scope: Projects sharing the pattern Participants: Affected project managers, PMO, subject matter experts


Retrospective Gates

Gate 1: Context Setting

Objective: Establish retrospective scope and participants

Actions:

  1. Define retrospective type and scope
  2. Identify participants
  3. Gather project data
  4. Schedule sessions

Context Template:

ElementValue
TypeProject/Portfolio/Thematic
Scope[Projects/Period]
Participants[List]
Data Sources[List]

Output: docs/pmo/{date}/retro-context.md


Gate 2: Data Collection

Objective: Gather objective data about performance

Actions:

  1. Collect final project metrics
  2. Gather variance explanations
  3. Document scope changes
  4. Compile risk/issue history

Data Points:

CategoryData
SchedulePlanned vs actual duration
CostBudget vs actual spend
ScopeBaseline vs final scope
QualityDefects, rework
RisksRisks that materialized
ChangesNumber and impact

Output: docs/pmo/{date}/retro-data.md


Gate 3: Reflection

Objective: Identify what worked, what didn't, and why

Actions:

  1. Conduct retrospective session
  2. Identify successes to repeat
  3. Identify failures to prevent
  4. Analyze root causes

Reflection Framework:

CategoryQuestion
What went well?What should we keep doing?
What didn't go well?What should we stop doing?
What was confusing?What needs clarification?
What was missing?What should we start doing?
What surprised us?What assumptions were wrong?

Output: docs/pmo/{date}/retro-reflection.md


Gate 4: Pattern Analysis

Objective: Identify patterns across projects/time

Actions:

  1. Compare with previous retrospectives
  2. Identify recurring themes
  3. Assess systemic vs isolated issues
  4. Prioritize improvement areas

Pattern Types:

TypeIndicatorAction
SystemicSame issue in 3+ projectsProcess change required
CapabilitySame team struggle repeatingTraining/hiring needed
ToolTool-related frictionTool improvement/replacement
CommunicationStakeholder issues recurringCommunication improvement
EstimationConsistent over/under estimationEstimation process improvement

Output: docs/pmo/{date}/retro-patterns.md


Gate 5: Action Planning

Objective: Create actionable improvement plan

Actions:

  1. Define specific improvements
  2. Assign owners
  3. Set timelines
  4. Define success criteria

Action Template:

ImprovementOwnerTimelineSuccess CriteriaStatus
[Improvement][Owner][Date][Criteria][Status]

Priority Framework:

Impact / EffortLow EffortHigh Effort
High ImpactDo FirstPlan Carefully
Low ImpactQuick WinsDon't Do

Output: docs/pmo/{date}/retro-actions.md


Gate 6: Knowledge Sharing

Objective: Distribute lessons to organization

Actions:

  1. Document lessons learned
  2. Update PMO knowledge base
  3. Present to relevant teams
  4. Update templates/processes

Knowledge Sharing Channels:

ChannelAudienceContent
PMO WikiAll PMsFull lessons
NewsletterOrganizationSummary
TrainingNew PMsIncorporated
TemplatesAll projectsUpdated
PlaybooksSpecific scenariosDetailed guidance

Output: docs/pmo/{date}/lessons-learned.md


Anti-Rationalization Table

See shared-patterns/anti-rationalization.md for universal anti-rationalizations.

Retrospective-Specific Anti-Rationalizations

RationalizationWhy It's WRONGRequired Action
"We're too busy for retrospectives"Learning gaps cost more than retrospective time.Schedule and conduct retrospective
"We know what went wrong"Assumptions miss root causes.Formal analysis required
"It was a unique situation"Patterns hide in "unique" situations.Document and compare
"People will be defensive"Blame-free culture enables learning.Focus on process, not people
"Lessons will be ignored anyway"Track action completion to ensure learning.Follow up on actions

Pressure Resistance

See shared-patterns/pressure-resistance.md for universal pressure scenarios.

Retrospective-Specific Pressures

Pressure TypeRequestAgent Response
"Skip retro, team already on next project""Learning before moving on prevents repeating mistakes. Conducting abbreviated retrospective."
"Don't document that failure""Failures are learning opportunities. Documenting with constructive framing."
"Just move on, it's in the past""Past informs future. Retrospective protects future projects."

Blocker Criteria - STOP and Report

ALWAYS pause and report blocker for:

SituationRequired Action
Key participants unavailableSTOP. Incomplete retrospective misses perspectives. Reschedule.
Data unavailableSTOP. Data-free retrospective is opinion session. Get data.
Blame culture emergingSTOP. Reset to blameless framework. Not about individuals.
No action ownershipSTOP. Lessons without owners become forgotten. Assign owners.

Cannot Be Overridden

The following requirements are NON-NEGOTIABLE:

RequirementCannot Override Because
Blameless approachBlame prevents learning. Fear prevents honesty.
Action owner assignmentUnowned actions are never completed.
Data-driven analysisOpinions without data lead to wrong conclusions.
Participant inclusionMissing perspectives create incomplete learning.
Follow-up trackingLessons without follow-up are forgotten.

If user insists on violating these:

  1. Escalate to orchestrator
  2. Do NOT proceed with incomplete retrospective
  3. Document the request and your refusal

Severity Calibration

When assessing retrospective findings:

SeverityCriteriaExamples
CRITICALSystemic issue affecting multiple projectsProcess failure causing 3+ project delays, compliance violation pattern
HIGHSignificant recurring problemSame estimation error in 2+ projects, repeated resource conflicts
MEDIUMNotable improvement opportunityCommunication gaps, tool friction, documentation quality
LOWMinor optimizationTemplate improvements, minor process tweaks

Document ALL severities. Prioritize action on CRITICAL and HIGH.


Output Format

Lessons Learned Report

# Lessons Learned - [Project/Period] - [Date]

## Context

| Element | Value |
|---------|-------|
| Scope | [Project(s)/Period] |
| Participants | [List] |
| Facilitator | [Name] |

## Performance Summary

| Metric | Target | Actual | Variance |
|--------|--------|--------|----------|
| Duration | X months | X months | +/- X% |
| Budget | $X | $X | +/- X% |
| Scope | X features | X features | +/- X |
| Quality | X defects | X defects | +/- X |

## What Went Well

| Success | Why It Worked | How to Repeat |
|---------|---------------|---------------|
| [Success] | [Root cause] | [Action to sustain] |

## What Didn't Go Well

| Issue | Root Cause | How to Prevent |
|-------|------------|----------------|
| [Issue] | [Root cause] | [Action to prevent] |

## Patterns Identified

| Pattern | Frequency | Systemic? | Action |
|---------|-----------|-----------|--------|
| [Pattern] | X occurrences | Yes/No | [Action] |

## Improvement Actions

| # | Improvement | Owner | Due | Status |
|---|-------------|-------|-----|--------|
| 1 | [Improvement] | [Owner] | [Date] | [Status] |

## Key Lessons (Top 5)

1. **[Lesson Title]:** [Description and application guidance]
2. **[Lesson Title]:** [Description and application guidance]
3. **[Lesson Title]:** [Description and application guidance]
4. **[Lesson Title]:** [Description and application guidance]
5. **[Lesson Title]:** [Description and application guidance]

## Knowledge Sharing Plan

| Audience | Channel | Date | Owner |
|----------|---------|------|-------|
| [Audience] | [Channel] | [Date] | [Owner] |

Execution Report

Base metrics per shared-patterns/execution-report.md:

MetricValue
Analysis DateYYYY-MM-DD
Scope[Project(s)/Period]
DurationXh Ym
ResultCOMPLETE/PARTIAL/BLOCKED

Retrospective-Specific Details

MetricValue
period_covered[Description]
projects_completedN
lessons_capturedN
process_improvementsN

When Retrospective Is Not Needed

ConditionVerification
Project was trivial (<1 week, <3 people)Verify scope and team size
No significant issues occurredConfirm no deviations from plan
Team explicitly requested skipWritten confirmation required
Previous recent retrospective covers patternsReference recent retro that applies

MUST: Full retrospective REQUIRED for the following conditions:

ConditionWhy Required
Any project completionLearning opportunity, even for successful projects
Any project terminationUnderstanding failure prevents repetition
Significant budget/schedule varianceRoot cause analysis prevents recurrence
Team or stakeholder conflictsRelationship and process lessons critical
Process improvement identifiedMust capture and act on improvement

MUST: When in doubt, conduct the retrospective. Skipped retrospectives become repeated failures.

平台分布

Codex

36.56%
按下载量换算76

Claude

32.33%
按下载量换算67

Cursor

16.8%
按下载量换算35

Gemini CLI

9.26%
按下载量换算19

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills