Token导航 LogoToken导航

Agent Skill · 研究检索

deep-productivity(深度生产力)

deep-productivity 用于记录任务执行中的错误、用户纠正、经验和能力缺口,适合在 Codex、Claude、Cursor、Gemini CLI 中希望让 Agent 持续沉淀问题、修正和最佳实践时使用。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。

deep-productivity 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 7,486。

总安装 18,863周安装 802GitHub Stars 公开资料未说明下载 6,608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:deep-productivity(深度生产力)
来源仓库:https://github.com/hexbee/hello-skills
仓库路径:skills/deep-productivity
安装命令:
npx skills add https://github.com/hexbee/hello-skills --skill deep-productivity
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/hexbee/hello-skills --skill deep-productivity

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。

通过三部分框架掌握深度工作:每天 1 小时的构建,与维护和恢复保持平衡。

通过 90 天的转型序列建立可持续的生产力:诊断焦点状态、定义愿景和反愿景、建立 10 年到每日的目标层次结构,并应用基于项目的学习来缩小技能差距 使用挑战等于技能原则,确定真正推进目标的杠杆移动任务,避免焦虑(挑战太高)和无聊(挑战太低) 平衡重点工作和结构化恢复时间(步行、阅读、做白日梦),以激活默认模式网络并产生突破性想法 包括用于确定您处于哪个阶段(基础前、基础、构建或维护)的决策树以及包含会话前、会话期间和会话后步骤的日常深度工作协议

来源仓库内路径:skills/deep-productivity

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/hexbee/hello-skills --skill deep-productivity;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

deep-productivity 怎么安装?

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

deep-productivity 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。

deep-productivity 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。

deep-productivity 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。

deep-productivity 用的人多吗?

公开数据是累计安装 18,863 次,近一周安装 802 次,下载 6,608 次。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。

deep-productivity 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
deep-productivity

深度生产力 · 本页

18,863—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

Deep Productivity Mastery

Transform your life with just 1 hour/day of focused building work. This skill applies the three types of work framework—Building, Maintenance, and Recovery—to help you create meaningful progress toward your goals.

When to Use This Skill

  • User wants to build something meaningful (business, creative work, personal project)
  • User struggles with focus, distractions, or unclear priorities
  • User has limited time (1 hour/day) and wants to maximize impact
  • User feels stuck in "doing mode" without real progress
  • User wants to design a sustainable productivity system

Core Framework: Three Types of Work

Before diving into workflows, understand the three types of work:

  1. Building - Intense bursts of deep work to bring a project to life (products, services, brands, skills)
  2. Maintenance - Consistent, often repetitive work to keep what you built alive (marketing, customer service, operations)
  3. Recovery - Rest, leisure, and lack of narrow focused stress that allows breakthrough ideas to form

Your goal: Build for 1 hour/day until you can pursue your vision full-time. Then transition to maintenance work.

Workflow: The 90-Day Deep Work Transformation

Follow this sequence to establish a sustainable deep work practice:

Step 1: Diagnose Your Current State

Assess where you are on the focus spectrum:

StateSymptomsRoot Cause
DistractedMultiple programs running, attention splitLack of clarity
AnxiousOverwhelmed by goals, paralyzedChallenge >> Skill
BoredRepetitive work, self-centered thoughtsChallenge << Skill
FlowChallenge ≈ Skill, time fliesOptimal state

Key insight: Humans can process ~50 bits of conscious info/second. Every distraction reduces your potential. You have ~125 billion bits total in a lifetime—use them wisely.

Step 2: Create Vision & Anti-Vision

Vision: Your ideal future (be specific, even if it seems irrational) Anti-Vision: What you don't want, where you'll end up without change

Exercise (15 minutes):

  1. Write out everything you DON'T want in your life—feel the discomfort
  2. Write out your ideal future—be petty, think big, ignore "rational" limits
  3. List all distractions standing between you and that future
  4. Identify people, apps, activities to temporarily remove

Rule: Success is less about discipline and more about removing distractions that make discipline difficult.

Step 3: Build Your Goal Hierarchy

Create a clear mental frame you can tap into anytime:

10-Year Goal → Where you want to be
    ↓
1-Year Goal → Milestone to reach this year
    ↓
1-Month Goal → What you must achieve this month
    ↓
1-Week Goal → Actionable tasks this week
    ↓
Lever-Moving Tasks → Daily 1-3 priorities that move the needle

Example:

  • 10-Year: Build a global digital education company
  • 1-Year: Launch MVP and get 1000 paid subscribers
  • 1-Month: Create minimum viable course product
  • 1-Week: Record 3 video lessons, set up payment system
  • Daily: Record 1 lesson (30 min), promote on social (30 min)

Step 4: Apply Project-Based Learning

Principle: Learn by building, not by consuming tutorials.

Why tutorials fail: Endless consumption creates endless options and overwhelm. Most information goes to waste because you can't apply it.

How to use project-based learning:

  1. Choose something to build that moves the needle toward your vision
  2. Brain dump everything you know about it
  3. Save 3-5 sources to emulate—study their structure
  4. Outline into sections, milestones, knowledge gaps
  5. Start building with what you know
  6. Let the project expose gaps—search for answers as needed

A project can be anything: Your health, a business, a skill, a creative work. A project is simply a structured way to achieve a goal.

Step 5: Design Your 1-Hour Deep Work Protocol

Structure your daily deep work block using the clarity-importance-urgency framework:

Before the session (2 min):

  • Review today's lever-moving task (from hierarchy)
  • Ask: "If I could only do one thing today, what would it be?"

During the session (50 min):

  • Single-task on that one thing
  • If blocked, break into smaller sub-goals
  • Apply challenge ≈ skill: if anxious, simplify; if bored, increase challenge

After the session (8 min):

  • Journal what you accomplished
  • Identify the next micro-step
  • Note any ideas that emerged (Default Mode Network will generate insights during rest)

Optimal timing: Morning, after filling your mind with education/ideas, before emptying with planning.

Step 6: Balance Work with Recovery

Critical: Your best ideas come during rest, not during work.

Daily structure (the routine of successful creatives):

  • Morning: Creation, output, focus (deep work block)
  • Afternoon: Learning, ideas, socialization (fill your mind)
  • Evening: Journaling, planning, meditation (empty your mind)

Recovery activities (not escapism):

  • Walking (especially without phone)
  • Reading fiction or philosophy
  • Daydreaming, imagining futures
  • Meaningful conversation
  • Solitude without distractions

The Default Mode Network: When you stop focusing externally, your brain connects regions for visual thinking and creativity. This is where breakthrough ideas emerge.

Decision Tree: What to Work On

What stage are you in?
├─ Pre-foundation (no clear vision)
│  └─ Focus: Vision & Anti-Vision exercise (Step 2)
│
├─ Foundation (vision set, no project)
│  └─ Focus: Goal hierarchy + First project definition (Steps 3-4)
│
├─ Building (active project in progress)
│  └─ Focus: Daily 1-hour protocol + Lever-moving tasks (Steps 5)
│
└─ Maintenance (project generating value)
    └─ Focus: Systematizing, delegating, starting next project

Quick Reference: The Deep Work Checklist

Before starting each session, confirm:

  • I know exactly what one task I'm working on
  • The challenge is just above my current skill level
  • I understand WHY this matters to my future
  • Distractions are removed (phone, tabs, notifications)
  • I have a hard stop time (60 or 90 minutes max)

After each session, confirm:

  • I made measurable progress on my project
  • I identified the next micro-step
  • I captured any ideas that emerged
  • I'm transitioning to recovery (not maintenance work)

Embracing Uncertainty

Expect to feel lost, overwhelmed, and like you don't know what you're doing. This is normal and signals growth.

  • Uncertainty is signal, not noise
  • Your brain filters reality based on what you're obsessed with
  • Invest in your portfolio of failures until you can afford to succeed
  • The "certain" life is the least rewarding

Key insight: Neurons that fire together, wire together. Being extreme about your goals accelerates neuroplasticity.

Resources

references/

  • focus-types.md - Detailed breakdown of Building, Maintenance, Recovery
  • flow-formula.md - The challenge ≈ skill equation explained
  • default-mode-network.md - How recovery generates creative insights

scripts/

  • assess_focus.py - Diagnose current focus state
  • generate_hierarchy.py - Create goal hierarchy from vision statement
  • track_levers.py - Track lever-moving tasks and progress

平台分布

Claude Code

29.99%
按下载量换算1,982

Gemini CLI

20.28%
按下载量换算1,340

Antigravity

17.87%
按下载量换算1,181

OpenCode

12.55%
按下载量换算829

windsurf

6.94%
按下载量换算459

Codex

3.01%
按下载量换算199

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