Token导航 LogoToken导航

Agent Skill · 研究检索

workflow-optimizer(工作流程优化器)

workflow-optimizer 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。

workflow-optimizer 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,权限为需要联网;许可证为 MIT,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 40,865。

总安装 1,247周安装 53GitHub Stars 14下载 437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
MIT

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:workflow-optimizer(工作流程优化器)
来源仓库:https://github.com/vladm3105/aidoc-flow-framework
仓库路径:skills/workflow-optimizer
安装命令:
npx skills add https://github.com/vladm3105/aidoc-flow-framework --skill workflow-optimizer
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/vladm3105/aidoc-flow-framework --skill workflow-optimizer

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。

用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中快速定位候选结果。支持基于关键词、任务场景或来源线索进行信息筛选,适用于优化场景。通过 npx skills add 命令从指定 GitHub 仓库安装,需确认权限与维护状态。

安装前建议核实是否会触发联网、命令执行或文件读写操作。可结合原始 README 进一步了解具体用法和功能边界。

来源仓库内路径:skills/workflow-optimizer

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/vladm3105/aidoc-flow-framework --skill workflo…;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

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

workflow-optimizer 怎么安装?

在终端执行 npx skills add https://github.com/vladm3105/aidoc-flow-framework --skill workflow-optimize…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

workflow-optimizer 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。

workflow-optimizer 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

workflow-optimizer 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

workflow-optimizer 用的人多吗?

公开数据是累计安装 1,247 次,近一周安装 53 次,下载 437 次,GitHub 14 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。

workflow-optimizer 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。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 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。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 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
workflow-optimizer

工作流程优化器 · 本页

1,24714需要联网MIT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

workflow-optimizer

Purpose

Guide users through the SDD workflow by determining current position, recommending next steps, identifying parallel work opportunities, and tracking progress.

Problem Solved: Users must manually determine next steps in the 15-layer SDD workflow, leading to workflow friction, missed dependencies, and inefficient sequencing.

Solution: Analyze completed artifacts, determine workflow position, and provide prioritized recommendations for next steps with clear rationale.

When to Use This Skill

Use workflow-optimizer when:

  • Completed an artifact and need guidance on next steps
  • Starting documentation and need workflow overview
  • Want to identify parallel work opportunities
  • Need progress report on documentation completion
  • Unsure which artifacts to create next

Do NOT use when:

  • Need skill recommendation for specific task (use skill-recommender)
  • Need project context (use context-analyzer)
  • Validating artifacts (use trace-check or quality-advisor)

Skill Inputs

InputTypeRequiredDescription
project_rootstringYesRoot path of project to analyze
completed_artifactstringNoID of just-completed artifact (e.g., PRD-00)
focus_areastringNoOptional filter: "core-workflow", "quality", "planning"

Skill Workflow

Step 1: Analyze Project State

Scan project to determine documentation status:

Artifact Discovery:

# Discover all artifacts
find {project_root}/docs -name "*.md" -o -name "*.yaml" -o -name "*.feature"

Status Extraction: Extract status from Document Control section:

  • Draft
  • In Review
  • Approved
  • Superseded
  • Deprecated

Project State Model:

project_state:
  scan_timestamp: 2025-11-29T14:30:00Z
  artifacts_by_type:
    BRD:
      total: 3
      approved: 2
      draft: 1
      latest: BRD-03
    PRD:
      total: 2
      approved: 1
      draft: 1
      latest: PRD-00
    EARS:
      total: 0
    # ... etc
  total_artifacts: 25
  approved_artifacts: 18
  draft_artifacts: 7

Step 2: Determine Workflow Position

Map artifacts to SDD workflow layers:

Layer Definition:

workflow_layers:
  layer_1:
    type: BRD
    name: Business Requirements
    prerequisite: null
    description: "Business objectives and stakeholder needs"

  layer_2:
    type: PRD
    name: Product Requirements
    prerequisite: BRD
    description: "Product features and user needs"

  layer_3:
    type: EARS
    name: Formal Requirements
    prerequisite: PRD
    description: "WHEN-THE-SHALL formal requirements"

  layer_4:
    type: BDD
    name: Behavior Tests
    prerequisite: EARS
    description: "Gherkin acceptance scenarios"

  layer_5:
    type: ADR
    name: Architecture Decisions
    prerequisite: BDD
    description: "Technical decision records"

  layer_6:
    type: SYS
    name: System Requirements
    prerequisite: ADR
    description: "Technical system specifications"

  layer_7:
    type: REQ
    name: Atomic Requirements
    prerequisite: SYS
    description: "Detailed atomic requirements"

  layer_8:
    type: IMPL
    name: Implementation Plan
    prerequisite: REQ
    optional: true
    description: "WHO/WHEN project management"

  layer_9:
    type: CTR
    name: Interface Contracts
    prerequisite: [IMPL, REQ]
    optional: true
    description: "API contracts and schemas"

  layer_10:
    type: SPEC
    name: Technical Specifications
    prerequisite: REQ
    description: "YAML implementation specs"

  layer_11:
    type: TASKS
    name: Implementation Tasks
    prerequisite: SPEC
    description: "Code generation task lists - flows to Code layer"

Position Calculation:

workflow_position:
  completed_layers: [1, 2]  # BRD, PRD done
  in_progress_layers: [3]   # EARS in progress
  blocked_layers: [4, 5, 6, 7, 10, 11]  # Waiting on prerequisites
  ready_layers: [3]         # Can start now
  optional_ready: [8, 9]    # Optional, prereqs met

  current_position:
    layer: 3
    type: EARS
    status: in_progress

  progress_percentage: 18%  # 2 of 11 layers complete

Step 3: Identify Required Next Steps

Determine mandatory next artifacts:

Dependency Analysis:

dependency_graph:
  PRD-00:
    completed: true
    downstream_required:
      - EARS (Layer 3) - formal requirements
      - BDD (Layer 4) - test scenarios
    downstream_optional:
      - IMPL (Layer 8) - if complex project

  EARS (to be created):
    upstream_required:
      - PRD-00 ✓ (completed)
    will_enable:
      - BDD (Layer 4)
      - ADR (Layer 5)

Next Steps Priority:

PriorityArtifactRationale
P0EARSRequired downstream from PRD, blocks BDD and ADR
P1BDDCan start once EARS begun, enables ADR
P2ADRRequires BDD completion

Step 4: Identify Parallel Opportunities

Find work that can proceed in parallel:

Parallelization Rules:

parallel_opportunities:
  rule_1:
    name: "EARS and BDD overlap"
    condition: "EARS in progress"
    parallel_work: "BDD scenarios for completed EARS"
    benefit: "Faster progress through testing layer"

  rule_2:
    name: "Multiple feature tracks"
    condition: "Multiple BRDs exist"
    parallel_work: "PRD for each BRD"
    benefit: "Parallel feature development"

  rule_3:
    name: "ADR independence"
    condition: "Technical decisions needed"
    parallel_work: "ADRs can be written in parallel"
    benefit: "Architecture decisions don't block each other"

Parallel Work Output:

parallel_opportunities:
  can_parallelize:
    - track: "Feature A"
      current: EARS-01
      parallel: "Start BDD-01 scenarios for EARS-01 requirements"

    - track: "Feature B"
      current: PRD-00 complete
      parallel: "Start EARS-02 while Feature A progresses"

  blocked_parallelization:
    - item: "SPEC creation"
      blocker: "REQ layer incomplete"
      unblock_by: "Complete REQ-01 through REQ-05"

Step 5: Calculate Progress Metrics

Generate progress report:

Progress Metrics:

progress_report:
  overall:
    layers_complete: 2
    total_layers: 12
    percentage: 17%

  by_layer:
    - layer: 1 (BRD)
      status: complete
      artifacts: 3
      approved: 2

    - layer: 2 (PRD)
      status: complete
      artifacts: 2
      approved: 1

    - layer: 3 (EARS)
      status: in_progress
      artifacts: 0
      target: 3 (based on PRD features)

    - layer: 4 (BDD)
      status: blocked
      blocker: "EARS incomplete"

  estimated_remaining:
    artifacts: 25
    layers: 10

  critical_path:
    - EARS (blocks BDD)
    - BDD (blocks ADR)
    - REQ (blocks SPEC)
    - SPEC (blocks TASKS)

Step 6: Generate Recommendations

Provide actionable next-step guidance:

Recommendation Format:

recommendations:
  context:
    completed_artifact: PRD-00
    workflow_position: Layer 2 complete
    progress: 17%

  next_steps:
    - priority: P0
      action: "Create EARS document"
      artifact_type: EARS
      skill: doc-ears
      rationale: "Required downstream from PRD-00. EARS formalizes product features into WHEN-THE-SHALL requirements."
      estimated_effort: "Medium (2-4 hours)"
      blocks: [BDD, ADR]

    - priority: P1
      action: "Start BDD scenarios"
      artifact_type: BDD
      skill: doc-bdd
      rationale: "Can begin once EARS started. Write scenarios for completed requirements."
      estimated_effort: "Medium (2-4 hours)"
      parallel_with: EARS

    - priority: P2
      action: "Consider ADR for key decisions"
      artifact_type: ADR
      skill: doc-adr
      rationale: "If architectural decisions needed, document early. Requires BDD completion."
      estimated_effort: "Low-Medium (1-3 hours)"
      condition: "After BDD progress"

  parallel_opportunities:
    - "Feature B: Start PRD-03 while Feature A progresses through EARS/BDD"
    - "Technical: Draft ADRs for known architecture decisions"

  blocked_items:
    - item: "SPEC creation"
      reason: "Requires REQ completion (Layer 7)"
      unblock_path: "Complete layers 3-7 first"

  workflow_guidance:
    current_focus: "EARS creation for PRD-00 features"
    short_term: "Complete EARS → BDD → ADR sequence"
    medium_term: "Progress through SYS → REQ → SPEC"

  progress_summary:
    completed: "BRD-01, BRD-02, BRD-03, PRD-01, PRD-00"
    in_progress: "None"
    next_milestone: "Complete Layer 3 (EARS)"
    overall: "17% complete (2/12 layers)"

Example Usage

Example 1: Post-PRD Guidance

User Request: "I just finished PRD-00, what should I do next?"

Workflow Recommendations:

completed: PRD-00
position: Layer 2 complete

next_steps:
  1. Create EARS-01 (P0)
     - Formalize PRD-00 features into WHEN-THE-SHALL requirements
     - Run: /skill doc-ears

  2. Start BDD scenarios (P1)
     - Can begin once EARS requirements defined
     - Run: /skill doc-bdd

parallel_opportunity:
  "If other BRDs exist, you can create their PRDs in parallel"

progress: "17% → 25% after EARS completion"

Example 2: Project Overview

User Request: "Where am I in the documentation workflow?"

Workflow Status:

project_status:
  completed_layers:
    - Layer 1 (BRD): 3 documents ✓
    - Layer 2 (PRD): 2 documents ✓

  in_progress:
    - Layer 3 (EARS): 0 documents, target 5

  blocked:
    - Layers 4-11: Waiting on upstream completion

  progress: 18%

  critical_path:
    EARS → BDD → ADR → SYS → REQ → SPEC → TASKS

  recommended_focus:
    "Complete EARS layer to unblock BDD and ADR"

Example 3: Parallel Work Identification

User Request: "What can I work on in parallel?"

Parallel Opportunities:

current_tracks:
  track_a:
    name: "Core Platform"
    position: EARS creation
    next: BDD scenarios

  track_b:
    name: "Partner Integration"
    position: BRD complete
    next: PRD creation (independent of Track A)

parallel_work:
  - "Track B PRD can proceed while Track A completes EARS"
  - "ADRs for known decisions can be drafted early"
  - "BDD scenarios can start once first EARS requirements defined"

sequential_requirements:
  - "SPEC requires REQ completion - no parallel path"
  - "TASKS requires SPEC - sequential"

Integration with Other Skills

IntegrationDescription
context-analyzerProvides artifact inventory and traceability data
skill-recommenderReceives workflow position for better skill suggestions
doc-flowCan be invoked by doc-flow for workflow orchestration
project-mngtShares planning concepts (MVP, phases)

Quality Gates

Definition of Done

  • Project state analyzed
  • Workflow position calculated
  • Next steps prioritized (P0, P1, P2)
  • Parallel opportunities identified
  • Progress metrics calculated
  • Actionable recommendations generated

Performance Targets

MetricTarget
Analysis latency<1s for 50 artifacts
Recommendation generation<500ms
Progress calculation<200ms

Traceability

Required Tags:

@prd: PRD.000.004
@adr: ADR-000

Upstream Sources

SourceTypeReference
PRD-00Product RequirementsPRD-00
ADR-000Architecture DecisionADR-000

Downstream Artifacts

ArtifactTypeReference
doc-flowSkill ConsumerWorkflow orchestration
project-mngtSkill ConsumerRelease planning integration

Version Information

Version: 1.0.0 Created: 2025-11-29 Status: Active Author: AI Dev Flow Framework Team

平台分布

Gemini CLI

25.56%
按下载量换算112

Claude Code

21.73%
按下载量换算95

OpenCode

19.17%
按下载量换算84

Antigravity

12.07%
按下载量换算53

Codex

7.91%
按下载量换算35

windsurf

3.68%
按下载量换算16

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

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。

继续浏览同类 Skills