PonytailPonytail v1.0.1

A structured task preset that routes work through the Ponytail efficiency ladder and its focused review, audit, debt, and help commands.一个结构化任务预设,把工作路由到 Ponytail 效率阶梯,以及它的 review、audit、debt、help 命令。

About this task任务介绍

Ponytail is a task preset for teams that want a built-in guardrail against over-engineering. It turns the shared preset-task drawer into a skill-first workflow where one command decides both the active Ponytail skill and, when relevant, the intensity level.

Why use ponytail

It pressures every task toward the shortest working path

Ponytail starts with YAGNI, then checks the standard library, native platform features, already-installed dependencies, one-liners, and only then allows fresh code.

It keeps repository scope explicit

Like other preset tasks, Ponytail keeps selected repositories visible and typed as read or write, so simplification never becomes an excuse for touching the wrong codebase.

It covers more than build mode

The same preset also exposes review, audit, debt, and help, so users can inspect over-engineering, scan a repo, track deferred shortcuts, or pull up a quick reference without switching to another preset.

How the ladder works

  1. Skip speculative work.
  2. Prefer the standard library.
  3. Prefer native platform features.
  4. Reuse already-installed dependencies.
  5. Prefer the one-line version.
  6. Only then write the minimum new code that works.

The ladder never overrides trust-boundary validation, data-loss prevention, security controls, or accessibility basics.

Modes explained

  • lite: Build what was asked, but mention the lazier alternative.
  • full: Default mode. Enforce the ladder and keep the diff short.
  • ultra: Push back on unnecessary scope and ship the harshest minimal version.

Auxiliary commands stay available through the same picker:

  • review: code or diff review for over-engineering.
  • audit: repo-wide complexity scan.
  • debt: ledger of ponytail: shortcuts.
  • help: quick reference card.

Best fit

  • Best for: implementation tasks that tend to attract boilerplate, dependency creep, or speculative abstractions.
  • Also good for: focused review passes, cleanup audits, and teams that want a faster default path in non-interactive automation.
  • Not for: work that genuinely needs broad architecture expansion, extensive stakeholder clarification, or a deliberately maximal solution.

FAQ

What do I need installed first?

The preset requires skills:ponytail and expects the full Ponytail bundle to be available, including ponytail-review, ponytail-audit, ponytail-debt, and ponytail-help.

How do I pick a mode?

Choose a command in the picker. full is the recommended starting point, while lite and ultra adjust how aggressively Ponytail applies the ladder.

Is Ponytail suitable for every task?

No. It is best when you want the simplest correct path. If the requirement truly needs a fuller build, Ponytail can still execute it, but it will keep challenging unneeded complexity.

Ponytail 是一个面向团队的任务预设,用来把“避免过度工程化”变成内置护栏。它把共享预设任务抽屉变成技能优先的工作流,一个命令同时决定激活哪个 Ponytail 技能,以及在需要时使用哪种强度等级。

为什么使用 ponytail

它会把每个任务都压向最短可行路径

Ponytail 先问 YAGNI,再检查标准库、原生平台能力、已安装依赖、一行写法,最后才允许新增代码。

它让仓库范围保持显式

和其他预设任务一样,Ponytail 会把被选中的仓库及其 read / write 权限清楚展示出来,简化并不意味着可以越界修改代码库。

它不只覆盖实现模式

同一个预设还提供 reviewauditdebthelp,所以用户可以在不切换预设的情况下做过度工程化审查、全仓扫描、延期项追踪或查看速查卡。

阶梯如何工作

  1. 跳过投机性需求。
  2. 优先标准库。
  3. 优先原生平台能力。
  4. 复用已安装依赖。
  5. 优先一行实现。
  6. 只有前面都不成立时,才编写最少的新代码。

这个阶梯不会覆盖信任边界校验、防止数据丢失的处理、安全控制或基础可访问性要求。

模式说明

  • lite:完成需求,同时指出更懒的替代方案。
  • full:默认模式。执行完整阶梯,保持最短 diff。
  • ultra:更强硬地质疑多余范围,并交付最小版本。

同一个命令选择器里还包含辅助命令:

  • review:针对过度工程化的代码或 diff 审查。
  • audit:仓库级复杂度扫描。
  • debt:收集 ponytail: 快捷注释形成账本。
  • help:查看速查卡。

最适合的场景

  • 最适合:容易长出样板代码、依赖膨胀或投机性抽象的实现任务。
  • 同样适合:聚焦型审查、清理审计,以及希望在非交互自动化中采用更快默认路径的团队。
  • 不适合:确实需要大规模架构扩展、需要大量澄清,或明确追求“最大化实现”的工作。

常见问题

使用前需要先安装什么?

该预设要求存在 skills:ponytail,并默认完整 Ponytail 套件都已可用,包括 ponytail-reviewponytail-auditponytail-debtponytail-help

我该怎么选模式?

直接在命令选择器里选择。full 是推荐起点,liteultra 则用来调整 Ponytail 应用阶梯的激进程度。

Ponytail 适合所有任务吗?

不是。它最适合你想要“最简单且正确路径”的场景。如果需求确实需要更完整的实现,Ponytail 也能执行,但它仍会持续质疑不必要的复杂度。