utility-pm-workflow-builder

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

PM Workflow Builder

This skill creates new multi-skill workflows for the pm-skills library. It produces a Workflow Implementation Packet - a draft

_workflows/<name>.md

file, a draft

commands/workflow-<name>.md

command, and the cross-cutting update checklist - in a staging area for review before promotion to canonical locations. The builder authors; it never executes a chain and it never promotes its own output.

该skill为pm-skills库创建新的多技能工作流。它会生成一个Workflow Implementation Packet——包括草稿

_workflows/<name>.md

文件、草稿

commands/workflow-<name>.md

命令，以及跨领域更新检查清单——存储在暂存区中，待审核后再推广到标准位置。此构建工具仅负责编写内容；它从不执行链，也从不自行推广输出结果。

When to Use

适用场景

You have an idea for a new multi-skill workflow ("I need a workflow for quarterly business reviews")
You know the skills you want to chain and want them captured durably ("chain competitive-analysis, experiment-results, pivot-decision")
A
```
/chain
```
run (orchestrator Mode B) proved reusable and its completion output suggested promoting the chain to a durable workflow; hand this skill the exact chain expression from that suggestion

你有一个新的多技能工作流想法（例如“我需要一个季度业务评审的工作流”）
你明确想要串联的skill，并希望将其持久化记录（例如“串联competitive-analysis、experiment-results、pivot-decision”）
某次
```
/chain
```
运行（编排器模式B）被证明可复用，且其完成输出建议将该链推广为持久工作流；请将该建议中的完整链表达式提供给此skill

When NOT to Use

不适用场景

To run a sequence once without authoring anything -> use the
```
/chain
```
command or
```
utility-pm-workflow-orchestrator
```
directly

To build or modify a single skill -> use

utility-pm-skill-builder

utility-pm-skill-validate

, or

utility-pm-skill-iterate

To customize an existing workflow -> edit its
```
_workflows/<name>.md
```
directly; the overlap analysis below will point you there when coverage is high

仅需运行一次序列而无需编写任何内容 -> 直接使用
```
/chain
```
命令或
```
utility-pm-workflow-orchestrator
```

构建或修改单个skill -> 使用

utility-pm-skill-builder

、

utility-pm-skill-validate

或

utility-pm-skill-iterate

自定义现有工作流 -> 直接编辑其
```
_workflows/<name>.md
```
文件；当重叠度较高时，下文的重叠分析会指向该文件

Instructions

操作步骤

Step 1: Understand the Idea

步骤1：理解想法

Accept the idea in any of three entry forms, all converging on the same downstream flow:

Problem-first: "I need a workflow for X." Identify the recurring PM process, who runs it, and the artifact trail it should leave.
Skills-first: a list of skills the user already wants to chain.
Chain-promotion: a literal chain expression (for example
```
measure-experiment-results -> discover-stakeholder-summary
```
), usually pasted from the orchestrator's completion suggestion, plus the context the run used.

If the idea is vague, ask ONE clarifying question (the recurring trigger and the final artifact), then proceed. Do not interrogate.

接受三种形式的输入，所有输入都会导向相同的后续流程：

问题导向：“我需要一个针对X的工作流。”确定重复出现的PM流程、执行者以及该流程应留下的产物轨迹。
Skill导向：用户已确定要串联的skill列表。
链推广：一个字面意义上的链表达式（例如
```
measure-experiment-results -> discover-stakeholder-summary
```
），通常从编排器的完成建议中复制而来，再加上该运行使用的上下文。

如果想法模糊，只需提出一个澄清问题（重复触发因素和最终产物），然后继续。不要过度询问。

Step 2: Overlap Analysis + Why Gate

步骤2：重叠分析 + Why Gate

Scan the

_workflows/

directory at run time for the current inventory (list every

*.md

except

README.md

). Never rely on a remembered count or a hardcoded list. Compare the idea against ALL existing workflows:

Name each overlapping workflow and what it covers.
Estimate coverage overlap honestly.

Kill gate (>70% coverage overlap): recommend, in order, (a) customizing the existing workflow, (b) adding a step to it, or (c) just using

/chain

for the occasional run. Do not proceed.

Why Gate (any meaningful overlap): ask for 2-3 specific scenarios where the existing workflows fail to produce what is needed. Do not pass the gate without that evidence or an explicit user override (record the override in the packet's Decision section).

运行时扫描

_workflows/

目录获取当前工作流清单（列出除

README.md

外的所有

*.md

文件）。切勿依赖记忆中的数量或硬编码列表。将想法与所有现有工作流进行比较：

列出每个重叠工作流及其覆盖范围。
如实估算覆盖重叠度。

终止门槛（重叠度>70%）：按顺序推荐(a)自定义现有工作流，(b)为其添加步骤，或(c)仅在偶尔运行时使用

/chain

。不再继续后续流程。

Why Gate（存在任何有意义的重叠）：要求提供2-3个现有工作流无法满足需求的具体场景。若无此类证据或用户明确覆盖（需在包的决策部分记录覆盖情况），则无法通过此门槛。

Step 3: Sequence Design

步骤3：序列设计

Select and order the steps with the user. Apply the chain-expression contract's validation rules at authoring time (

skills/utility-pm-workflow-orchestrator/references/PARSE-CONTRACT.md

, Mode B Chain Expression Contract):

Every step must resolve to an installed skill at
```
skills/<name>/SKILL.md
```
by EXACT name. Never approximate or auto-correct; on a miss, refuse the step and OFFER the closest real names (an offer, never a substitution).
Category 1 content skills only. Refuse as steps, each with a one-line explanation: Tier-3 maintenance skills (
```
utility-pm-skill-*
```
,
```
utility-pm-release-conductor
```
,
```
utility-pm-changelog-curator
```
,
```
utility-update-pm-skills
```
,
```
foundation-prioritized-action-plan
```
), dispatch skills that fan out to sub-agents (
```
utility-pm-critic
```
,
```
utility-pm-workflow-orchestrator
```
, and peers), and existing workflows (a workflow never nests a workflow). Self-reference is impossible by construction: this builder is itself Tier-3.
For EACH step, author the handoff: what the step consumes from the prior step's output, and what it must produce for the next. This authored handoff guidance is exactly what distinguishes a durable workflow from an ad-hoc chain; a packet without real handoffs is not ready.

与用户共同选择并排序步骤。在编写时应用链表达式契约的验证规则（

skills/utility-pm-workflow-orchestrator/references/PARSE-CONTRACT.md

，模式B链表达式契约）：

每个步骤必须通过精确名称映射到已安装的skill（路径为
```
skills/<name>/SKILL.md
```
）。切勿近似或自动纠正；若匹配失败，拒绝该步骤并提供最接近的真实名称（仅提供，绝不替换）。
仅允许第一类内容skill。拒绝以下类型的步骤，并给出一行解释：三级维护skill（
```
utility-pm-skill-*
```
、
```
utility-pm-release-conductor
```
、
```
utility-pm-changelog-curator
```
、
```
utility-update-pm-skills
```
、
```
foundation-prioritized-action-plan
```
）、可派生子Agent的调度skill（
```
utility-pm-critic
```
、
```
utility-pm-workflow-orchestrator
```
及同类工具），以及现有工作流（工作流从不嵌套工作流）。从结构上看，自我引用是不可能的：此构建工具本身属于三级工具。
为每个步骤编写交接说明：该步骤从上一步输出中接收什么内容，以及必须为下一步生成什么内容。这种编写好的交接指导正是持久工作流与临时链的区别所在；没有真实交接说明的包不算完成。

Step 4: Generate the Packet

步骤4：生成工作流包

Write the complete Workflow Implementation Packet to

_staging/workflows/<name>/

(gitignored; the same review model as the skill builder). Never write to canonical locations.

_staging/workflows/<name>/
├── workflow.md      <- draft _workflows/<name>.md
├── command.md       <- draft commands/workflow-<name>.md
└── PACKET.md        <- decision, overlap analysis, checklist, promotion steps

Use

references/TEMPLATE.md

as the packet format. The workflow draft must carry the full section inventory the current

_workflows/*.md

files share (see the TEMPLATE's embedded skeleton); the command draft mirrors the existing

commands/workflow-*.md

shape, including one literal

skills/<name>/SKILL.md

path per step and the

$ARGUMENTS

footer. Derive a linear mermaid context-flow diagram from the final sequence (step names as nodes, execution order as edges).

将完整的Workflow Implementation Packet写入

_staging/workflows/<name>/

（已被git忽略；与skill构建工具采用相同的审核模式）。切勿写入标准位置。

_staging/workflows/<name>/
├── workflow.md      <- 草稿 _workflows/<name>.md
├── command.md       <- 草稿 commands/workflow-<name>.md
└── PACKET.md        <- 决策、重叠分析、检查清单、推广步骤

使用

references/TEMPLATE.md

作为包的格式。工作流草稿必须包含当前所有

_workflows/*.md

文件共有的完整章节清单（参考模板中的嵌入框架）；命令草稿镜像现有

commands/workflow-*.md

的结构，包括每个步骤对应的

skills/<name>/SKILL.md

路径，以及

$ARGUMENTS

页脚。根据最终序列生成线性的mermaid上下文流程图（步骤名称为节点，执行顺序为边）。

Step 5: Review + Promotion Guidance

步骤5：审核 + 推广指导

Present the packet and stop. The builder itself never promotes. On approval, the user (or a follow-up session) moves the drafts to their canonical paths and works the packet's Cross-Cutting Checklist, which names every count and documentation surface that adding a workflow trips, including the validator-blind

.github/workflows/release.yml

release-note template. The builder never edits those cross-cutting files itself.

展示工作流包后停止操作。构建工具本身从不执行推广操作。获得批准后，用户（或后续会话）将草稿移至标准路径，并完成包中的跨领域检查清单，该清单列出了添加工作流时需要更新的所有计数和文档位置，包括验证器未覆盖的

.github/workflows/release.yml

发布说明模板。构建工具从不自行编辑这些跨领域文件。

Output Format

输出格式

The packet follows

references/TEMPLATE.md

Decision - recommendation plus Why Gate evidence (or the recorded override)
Overlap Analysis - every overlapping workflow, what it covers, the gap
Workflow Draft - complete
```
_workflows/<name>.md
```
content
Command Draft - complete
```
commands/workflow-<name>.md
```
content
Cross-Cutting Checklist - every surface to update at promotion, with its enforcing validator (or "validator-blind" where none exists)
Promotion Steps - the move-verify-validate sequence

See

references/EXAMPLE.md

for one complete worked packet.

工作流包遵循

references/TEMPLATE.md

的结构：

决策 - 建议及Why Gate证据（或记录的覆盖情况）
重叠分析 - 所有重叠工作流、其覆盖范围及差距
工作流草稿 - 完整的
```
_workflows/<name>.md
```
内容
命令草稿 - 完整的
```
commands/workflow-<name>.md
```
内容
跨领域检查清单 - 推广时需要更新的所有位置，以及对应的强制执行验证器（若无则标注“验证器未覆盖”）
推广步骤 - 移动-验证-确认的流程

完整示例可参考

references/EXAMPLE.md

。

Quality Checklist

质量检查清单

Before presenting the packet, verify:

Overlap analysis scanned the live
```
_workflows/
```
directory (no remembered list, no hardcoded count)
Why Gate evidence is specific (named scenarios) or an explicit override is recorded
Every step resolves to an installed skill by exact name
No Tier-3 maintenance skill, dispatch skill, or workflow appears as a step
Every step has an authored handoff (consumes / produces), not just a skill name
The workflow draft carries every section in the TEMPLATE's skeleton (metadata table, When to Use with Do NOT use, per-step blocks, context-flow diagram, tips, quality checklist, see also)
The command draft names each step's
```
skills/<name>/SKILL.md
```
path and ends with the
```
$ARGUMENTS
```
footer
The mermaid diagram matches the final sequence exactly
Everything written under
```
_staging/workflows/<name>/
```
; no canonical path touched
The Cross-Cutting Checklist is reproduced in full, including the
```
.github/workflows/release.yml
```
release-note row

展示工作流包前，请验证：

重叠分析扫描了实时的
```
_workflows/
```
目录（未使用记忆列表或硬编码数量）
Why Gate证据具体（已命名场景）或已记录明确的覆盖情况
每个步骤都通过精确名称映射到已安装的skill
步骤中未出现三级维护skill、调度skill或工作流
每个步骤都有编写好的交接说明（接收/生成内容），而非仅skill名称
工作流草稿包含模板框架中的所有章节（元数据表、适用与不适用场景、分步模块、上下文流程图、提示、质量检查清单、相关链接）
命令草稿列出了每个步骤的
```
skills/<name>/SKILL.md
```
路径，并以
```
$ARGUMENTS
```
页脚结尾
mermaid流程图与最终序列完全匹配
所有内容均写入
```
_staging/workflows/<name>/
```
；未触及标准路径
完整复制了跨领域检查清单，包括
```
.github/workflows/release.yml
```
发布说明行

Reference Files

参考文件

Packet template:
```
references/TEMPLATE.md
```
Worked example (quarterly business review):
```
references/EXAMPLE.md
```

Chain-expression contract applied at authoring time:

skills/utility-pm-workflow-orchestrator/references/PARSE-CONTRACT.md

Workflow exemplars:

_workflows/sprint-planning.md

_workflows/customer-discovery.md

Command exemplar:
```
commands/workflow-sprint-planning.md
```

Sibling builder for single skills:

skills/utility-pm-skill-builder/SKILL.md

包模板：
```
references/TEMPLATE.md
```
示例（季度业务评审）：
```
references/EXAMPLE.md
```

编写时应用的链表达式契约：

skills/utility-pm-workflow-orchestrator/references/PARSE-CONTRACT.md

工作流示例：

_workflows/sprint-planning.md

、

_workflows/customer-discovery.md

命令示例：
```
commands/workflow-sprint-planning.md
```

单个skill构建工具：

skills/utility-pm-skill-builder/SKILL.md