Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问clear审计未展示

writing-plans写作计划

Agent Skill

用于辅助文档、README、Markdown、说明文和内容稿件的整理与改写。它适合让 Agent 提炼结构、补齐章节、统一术语、检查链接或把零散材料整理成可读文档。使用时应保留项目已有事实、命令和路径,不要把未确认的信息写成确定结论;涉及对外文案时,还需要控制语气,避免过度营销或夸大能力。

总安装

196

周安装

8

GitHub Stars

40,306

下载量

63
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

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:writing-plans(写作计划)
来源仓库:https://github.com/ian-pascoe/dotfiles
仓库路径:skills/writing-plans
安装命令:
npx skills add ian-pascoe/dotfiles --skill "writing-plans"
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。

AgentSkills.tonpx skills
npx skills add ian-pascoe/dotfiles --skill "writing-plans"

简介

用于辅助文档、README、Markdown 和内容稿件的整理与改写。

  • 适合提炼结构、补齐章节、统一术语或检查链接,提升内容可读性。
  • 使用时需保留项目事实和路径,避免将未确认信息写成确定结论。
  • 涉及对外文案时应控制语气,防止过度营销或夸大能力。
  • 安装前请确认来源仓库和权限范围,确保维护状态正常。

SKILL.md

name
writing-plans
description
Use when creating, editing, or reviewing implementation plans. Covers plan structure, version headers, task phases, acceptance criteria, and checkpoint format.

Writing Plans

Overview

Plans are step-by-step implementation guides that enable executor agents to complete work autonomously. A plan breaks complex features into ordered, atomic tasks with clear acceptance criteria.

When to Use

  • Creating a new implementation plan
  • Editing or updating an existing plan
  • Reviewing plan quality before execution
  • Adding checkpoints for session handoffs
  • Understanding plan versioning rules

Don't use for: one-off tasks (use TodoWrite), project documentation (use AGENTS.md), or design decisions (use architect agent).

File Location

.agents/plans/<name>.md     # Project-local plans

Plan names should be kebab-case, matching the feature: user-avatar-upload.md, auth-refresh-tokens.md.

Quick Reference

ElementFormatRequired
Version Header**Version**: 1.0 + timestamp, agent, statusYes
Task IDX.Y (phase.task)Yes
File Path**File**: \path/to/file.ts\``Yes (code tasks)
Done WhenCheckbox list of criteriaYes
Phases### Phase N: NameYes
CheckpointSession, Completed, In Progress, Notes, BlockersWhen stopping
Version BumpWhen
+0.1Task status, add/remove task, phase completion
+1.0Major restructure

Version Header Format

Every plan MUST include a version header:

# Plan: [Feature Name]

**Version**: 1.0
**Last Updated**: 2024-01-15T10:00:00Z
**Last Agent**: planner
**Status**: Draft
**Complexity**: Low | Medium | High
**Tasks**: [N]

Version Incrementing

Change TypeVersion BumpExample
Initial creation1.0(new file)
Task status update+0.11.0 → 1.1
Add/remove task+0.11.1 → 1.2
Phase completion+0.11.2 → 1.3
Major restructure+1.01.3 → 2.0

Status Values

  • Draft: Plan created, not yet started
  • In Progress: Execution underway
  • Paused: Stopped due to blocker
  • Complete: All tasks finished

Plan Structure Template

# Plan: [Feature Name]

**Version**: 1.0
**Last Updated**: [ISO timestamp]
**Last Agent**: planner
**Status**: Draft
**Complexity**: Low | Medium | High
**Tasks**: [N]

## Overview

[1-2 sentences describing the goal and scope]

## Tasks

### Phase 1: [Phase Name]

#### 1.1 [Task Name]

**File**: `path/to/file.ts`

[What to do - describe the change, not how to implement]

**Done when**:

- [ ] [Criterion 1]
- [ ] [Criterion 2]

#### 1.2 [Task Name]

**File**: `path/to/file.ts`

[Continue pattern...]

**Done when**:

- [ ] [Criterion 1]

### Phase 2: [Phase Name]

#### 2.1 [Task Name]

[Continue pattern...]

## Testing

- [ ] [Test category]: [What to verify]
- [ ] [Test category]: [What to verify]

## Risks

| Risk       | Mitigation          |
| ---------- | ------------------- |
| [Risk 1]   | [How to handle]     |
| [Risk 2]   | [How to handle]     |

Task Format

Each task must include:

  1. Numbered ID: X.Y format (phase.task)
  2. File: Target file path (required for code tasks)
  3. Description: What to do (not how)
  4. Done when: Checklist of acceptance criteria

Good Task Example

#### 2.1 Add Avatar Upload Route

**File**: `src/routes/users.ts`

Create POST /users/:id/avatar endpoint that accepts multipart form data.

**Done when**:

- [ ] Route accepts multipart/form-data
- [ ] Validates user owns profile or is admin
- [ ] Returns new avatar URL on success
- [ ] Returns 400 on invalid file type

Bad Task Example

#### 2.1 Implement avatar stuff

Add the avatar upload feature to the backend.

Problems: No file path, vague description, no acceptance criteria.

Checkpoint Format

When stopping mid-plan, add a checkpoint section:

## Checkpoint

**Session**: 2024-01-15T16:45:00Z
**Completed**: Tasks 1.1-1.4, 2.1-2.2
**In Progress**: Task 2.3 (started, ~50% done)
**Notes**: Using jose library per architect recommendation
**Blockers**: Waiting for API key from external team

Place checkpoint after the version header, before Overview.

Common Mistakes

MistakeFix
No file paths on tasksAlways include **File**: path/to/file.ts
Mega-tasks (multi-session)Split into atomic, single-session tasks
Missing acceptance criteriaEvery task needs **Done when** checklist
Describing HOW not WHATTask describes outcome, executor decides how
Unordered dependenciesTasks that block others must come first
No version headerAlways include version, timestamp, status
Skipping phasesGroup related tasks into logical phases
Vague criteria like "works"Specific, verifiable conditions

Validation Checklist

  • [ ] Version header includes all required fields
  • [ ] Each task has a numbered ID (X.Y format)
  • [ ] Each code task has a **File**: field
  • [ ] Each task has **Done when**: with checkboxes
  • [ ] Tasks are atomic (completable in one session)
  • [ ] Dependencies are ordered correctly
  • [ ] Phases group related work logically
  • [ ] Testing section lists verification approach
  • [ ] Risks identified with mitigations

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

windsurf

26.42%
按下载量换算17

OpenCode

23.06%
按下载量换算15

Codex

15.88%
按下载量换算10

Claude Code

13.82%
按下载量换算9

Antigravity

8.69%
按下载量换算5

Gemini CLI

3.91%
按下载量换算2

安全审计

暂无安全审计结果可展示。

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。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills