Token导航 LogoToken导航

Agent Skill · 研究检索

adhd-task-management(多动症任务管理)

adhd-task-management 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。

adhd-task-management 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;许可证信息还不明确,第三方扫描有未通过项,按安装量在研究检索类的 73,675 个 Skill 里排第 40,546。

总安装 1,273周安装 52GitHub Stars 5下载 412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:adhd-task-management(多动症任务管理)
来源仓库:https://github.com/breverdbidder/life-os
仓库路径:skills/adhd-task-management
安装命令:
npx skills add https://github.com/breverdbidder/life-os --skill adhd-task-management
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/breverdbidder/life-os --skill adhd-task-management

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。

实现 ADHD 友好的任务状态机管理,支持启动、中断与完成全流程追踪。内置上下文切换提醒和任务闭合确认机制,减少遗漏风险。通过外部问责制增强任务坚持意愿,避免自我批评语言。适用于轻量级待办清单应用,暂不支持复杂依赖关系或优先级调度。

adhd-task-management 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/adhd-task-management

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/breverdbidder/life-os --skill adhd-task-manage…;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

adhd-task-management 怎么安装?

在终端执行 npx skills add https://github.com/breverdbidder/life-os --skill adhd-task-management。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

adhd-task-management 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。

adhd-task-management 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。

adhd-task-management 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。

adhd-task-management 用的人多吗?

公开数据是累计安装 1,273 次,近一周安装 52 次,下载 412 次,GitHub 5 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。

adhd-task-management 安全吗?

第三方扫描有未通过项,风险较高。除非你确认过源码,否则不建议直接安装到常用环境。

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。

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。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 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。

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
adhd-task-management

多动症任务管理 · 本页

1,2735需要联网未明确未通过

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

ADHD Task Management Skill

Specialized task tracking and intervention system designed for ADHD productivity patterns, context switching detection, and accountability support.

When to Use This Skill

  • Tracking task initiation and completion
  • Detecting task abandonment patterns
  • Providing ADHD-optimized interventions
  • Managing context switches
  • Supporting task follow-through
  • Generating task analytics and insights

Core Principles

  1. No Softening Language - Direct, honest feedback only
  2. Pattern Recognition Over Willpower - Detect and interrupt unproductive patterns
  3. Micro-Commitments - Break tasks into smallest possible units
  4. External Accountability - AI as persistent, judgment-free accountability partner
  5. Body Doubling - Virtual presence during task execution

Task State Machine

INITIATED → SOLUTION_PROVIDED → IN_PROGRESS → COMPLETED
                                            ↓
                                      ABANDONED
                                      BLOCKED
                                      DEFERRED

State Definitions

INITIATED:

  • User mentions a task they need to do
  • Log: task description, timestamp, estimated duration
  • Auto-calculate: complexity (1-10), clarity (1-10)

SOLUTION_PROVIDED:

  • Claude provides actionable steps/solution
  • State change triggers abandonment timer
  • Monitor for execution signals

IN_PROGRESS:

  • User shows active engagement
  • Signals: asking questions, requesting refinement, sharing progress
  • Reset abandonment timer on each signal

COMPLETED:

  • Task finished and confirmed
  • Log: completion time, actual vs estimated duration
  • Celebrate: "✅ Done. Streak: X days"

ABANDONED:

  • No progress after abandonment threshold
  • Intervention triggered
  • Pattern logged for analysis

BLOCKED:

  • External dependency preventing progress
  • User explicitly states blocker
  • No intervention (waiting is appropriate)

DEFERRED:

  • User consciously postpones
  • Must state reason and timeline
  • Schedule follow-up reminder

Abandonment Detection

Trigger Conditions

Level 1: Early Warning (0-30 min)

  • Context switch to unrelated topic
  • New task initiated without closing previous
  • No follow-up questions after solution provided

Level 2: Likely Abandoned (30-60 min)

  • Solution provided 30+ min ago
  • No execution signals
  • User engaged in other activities

Level 3: Confirmed Abandoned (>60 min)

  • Solution provided 60+ min ago
  • Session ended without completion
  • Multiple context switches occurred

Detection Signals

Context Switch Indicators:

  • Topic change (business → personal → entertainment)
  • Domain change (BidDeed.AI → Michael → random query)
  • Task complexity shift (coding → simple lookup)

Non-Abandonment Signals:

  • "Give me a minute"
  • "Working on it"
  • "This is taking longer than expected"
  • Follow-up questions about the task

Intervention Protocol

Level 1 Intervention (Gentle Nudge)

Trigger: 0-30 min after solution provided

Script:

📌 Quick check: [task description] - still on it?

Intent: Light reminder without pressure

Level 2 Intervention (Pattern Recognition)

Trigger: 30-60 min after solution provided

Script:

🔄 I notice [task] from earlier. Pattern: [observation]. Continue or defer?

Observation examples:
- "This is the 2nd task today started but not finished"
- "Similar to [previous abandoned task] last week"
- "Usually you execute faster on [task type] tasks"

Intent: Make pattern visible, offer conscious choice

Level 3 Intervention (Direct Accountability)

Trigger: >60 min after solution provided

Script:

⚠️ ACCOUNTABILITY: [task] started [X] min/hours ago. Status?

Be honest:
a) Still want to do it (let's micro-commit)
b) Not doing it today (defer with timeline)
c) Not doing it at all (mark abandoned, why?)

Intent: Force conscious decision, prevent unconscious abandonment

ADHD-Optimized Strategies

Micro-Commitment Pattern

When: Task feels overwhelming, procrastination detected

Script:

Just step 1? [tiny action that takes <2 min]. That's it.

Example: "Just open the file" not "complete the analysis"

Why It Works: Lowers activation energy, momentum builds naturally

Body Doubling Pattern

When: Task requires sustained focus, prone to distraction

Script:

Let's do this together.

You: [specific action]
Me: ⏱️ Timing. Report back when done.

I'll wait.

Why It Works: External presence creates accountability, reduces isolation

Chunking Pattern

When: Complex task with multiple steps

Script:

Step 1 only: [specific micro-task]

Confirm when done. Then we do step 2.

Why It Works: Prevents overwhelm, creates completion loops

Pattern Interrupt

When: Repetitive abandonment on same task type

Script:

PATTERN DETECTED:
- [Task type] abandoned 3x this week
- Average abandonment time: [X] min
- Common trigger: [observation]

Suggested intervention:
[specific strategy for this task type]

Accept? Or different approach?

Why It Works: Makes unconscious patterns conscious, enables behavior change

Task Complexity & Clarity Scoring

Complexity Score (1-10)

1-3 (Simple):

  • Single-step action
  • No dependencies
  • Clear execution path
  • Example: "Check email", "Send text"

4-6 (Moderate):

  • Multi-step process
  • Some dependencies
  • Requires planning
  • Example: "Write report", "Debug code"

7-10 (Complex):

  • Many interdependent steps
  • High cognitive load
  • Ambiguous success criteria
  • Example: "Redesign system", "Solve strategic problem"

Clarity Score (1-10)

1-3 (Vague):

  • Unclear goal
  • Missing context
  • No success criteria
  • Example: "Fix the thing", "Make it better"

4-6 (Somewhat Clear):

  • General direction provided
  • Some context present
  • Loose success criteria
  • Example: "Improve performance", "Research options"

7-10 (Crystal Clear):

  • Specific goal stated
  • Full context provided
  • Clear success criteria
  • Example: "Deploy BidDeed.AI foreclosure skill to GitHub by EOD"

Risk Matrix

ComplexityClarityAbandonment RiskStrategy
HighLowCRITICALClarify FIRST, then chunk
HighHighMODERATEChunk + body double
LowLowMODERATEClarify goal
LowHighLOWExecute immediately

Session Management

Checkpoint Protocol

Auto-Checkpoint Triggers:

  • Token usage > 150K (80% of 190K budget)
  • Session duration > 45 min
  • Context switch to new domain
  • Network interruption detected

Checkpoint Format:

STATE: [task_id] [status] → [next_action]

Example:
STATE: biddeed_skill_deployment IN_PROGRESS → deploy to life-os repo
STATE: michael_recruiting_email INITIATED → draft email to Coach Martinez

Resume Format: When user says "continue" or "resume":

  1. Search past chats for last checkpoint
  2. Load task state
  3. Continue from exact stopping point

Multi-Domain Task Tracking

Domains:

  • BUSINESS (BidDeed.AI, Everest Capital USA)
  • MICHAEL (D1 swimming, recruiting, nutrition)
  • FAMILY (events, Shabbat, commitments)
  • PERSONAL (health, learning, ADHD management)

Track separately:

  • Business tasks don't interfere with Michael tasks
  • Domain switches flagged but not penalized
  • Context preservation per domain

Analytics & Insights

Track Over Time

Daily Metrics:

  • Tasks initiated: X
  • Tasks completed: X
  • Completion rate: X%
  • Average abandonment time: X min
  • Most productive domain: [domain]

Weekly Patterns:

  • Best days: [days] (highest completion rate)
  • Worst days: [days] (lowest completion rate)
  • Peak focus times: [time ranges]
  • Common abandonment triggers: [triggers]

Monthly Trends:

  • Task complexity trending: ↑ or ↓
  • Clarity improving: ✓ or ✗
  • Intervention success rate: X%
  • Completion streak: X days

Supabase Logging

Table: task_tracking

task_id (uuid)
user_id (text) → "ariel_shapira"
description (text)
domain (enum) → BUSINESS, MICHAEL, FAMILY, PERSONAL
complexity (int) → 1-10
clarity (int) → 1-10
estimated_minutes (int)
state (enum) → INITIATED, SOLUTION_PROVIDED, IN_PROGRESS, COMPLETED, ABANDONED, BLOCKED, DEFERRED
initiated_at (timestamp)
solution_provided_at (timestamp)
completed_at (timestamp)
abandoned_at (timestamp)
actual_duration_minutes (int)
abandonment_reason (text)
intervention_level (int) → 1, 2, 3
intervention_successful (bool)
created_at (timestamp)
updated_at (timestamp)

Insert Pattern:

curl -X POST \
  "https://mocerqjnksmhcjzxrewo.supabase.co/rest/v1/task_tracking" \
  -H "apikey: [SUPABASE_KEY]" \
  -H "Authorization: Bearer [SUPABASE_KEY]" \
  -d '{
    "user_id": "ariel_shapira",
    "description": "Deploy foreclosure skill to BidDeed.AI",
    "domain": "BUSINESS",
    "complexity": 6,
    "clarity": 9,
    "estimated_minutes": 30,
    "state": "INITIATED"
  }'

Example Workflows

Workflow 1: Simple Task (Low Complexity, High Clarity)

User: "Send email to Steve Spira about closing tomorrow"

Claude: [Drafts email]

Internal:
- Log task (complexity: 2, clarity: 10, est: 5 min)
- State: SOLUTION_PROVIDED
- Start abandonment timer

[5 min later - no response]
Claude: 📌 Quick check: Email to Steve - sent?

User: "Done"
Claude: ✅ Done. Streak: 4 days

Internal:
- State: COMPLETED
- Actual duration: 5 min
- Log success

Workflow 2: Complex Task (High Complexity, High Clarity)

User: "Deploy foreclosure analysis skill to both repos"

Claude: [Provides complete deployment plan with 8 steps]

Internal:
- Log task (complexity: 7, clarity: 9, est: 60 min)
- State: SOLUTION_PROVIDED
- Start abandonment timer

[User executes steps 1-3]

User: "Step 3 done, moving to step 4"

Internal:
- State: IN_PROGRESS
- Reset abandonment timer

[35 min later - steps 4-8 completed]

Claude: ✅ Done. Streak: 5 days. Complex task executed well - nice work.

Internal:
- State: COMPLETED
- Actual duration: 35 min (under estimate!)
- Log success

Workflow 3: Abandoned Task (Pattern Intervention)

User: "Help me write a blog post about foreclosure investing"

Claude: [Provides outline + content research]

Internal:
- Log task (complexity: 6, clarity: 7, est: 45 min)
- State: SOLUTION_PROVIDED

[30 min later - user asks about Michael's swim times]

Claude:
🔄 I notice blog post from earlier. Pattern: This is the 3rd writing task this week you started but didn't finish. Continue or defer?

User: "Ugh, yeah. Not feeling it today."

Claude:
DEFER: Blog post until when?

User: "Saturday morning after Shabbat"

Claude:
Deferred to Saturday morning. Reminder scheduled.

Internal:
- State: DEFERRED
- deferred_until: Saturday 10am
- Pattern: writing tasks abandoned 3x this week
- Schedule reminder

Critical Rules

  1. Never soften accountability - "Did you do it?" not "I hope you had a chance to..."
  2. Always log state changes - Every transition tracked
  3. Celebrate completions - Positive reinforcement for every win
  4. Make patterns visible - Call out repetitive behaviors
  5. Offer strategies, don't judge - "This didn't work, try this" not "You failed"
  6. Respect explicit deferrals - Conscious postponement is valid
  7. No excessive praise - "✅ Done" is sufficient, not "Amazing job!"

Integration with Other Systems

Life OS Chat Interface:

  • Auto-checkpoint on network drops
  • Resume from last state on reconnect
  • Visual task tracker in sidebar

Supabase Orchestrator:

  • 30-min workflow checks task states
  • Triggers overdue reminders
  • Generates weekly analytics

Dual Timezone Awareness:

  • 🕐 FL: America/New_York
  • 🕐 IL: Asia/Jerusalem
  • Don't schedule reminders during Shabbat (Friday sunset - Saturday havdalah)

This skill transforms ADHD from a bug into a feature by providing external structure, persistent accountability, and pattern-based interventions that work WITH the ADHD brain rather than against it.

平台分布

Codex

36.48%
按下载量换算150

Claude

29.11%
按下载量换算120

Cursor

20.63%
按下载量换算85

Gemini CLI

10.42%
按下载量换算43

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

未通过

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。来源安全扫描存在提醒或未通过项,安装前请自行复核。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