Token导航 LogoToken导航TokenDH.com
研究检索执行命令github未标认证来源可访问许可证需确认审计通过

engineer-plan工程师计划

Agent Skill

engineer-plan 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。

总安装

196

周安装

8

GitHub Stars

1

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

unknown

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:engineer-plan(工程师计划)
来源仓库:https://github.com/elliottrjacobs/bench-skills
仓库路径:skills/engineer-plan
安装命令:
npx skills add https://github.com/elliottrjacobs/bench-skills --skill engineer-plan
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/elliottrjacobs/bench-skills --skill engineer-plan

简介

工程师计划用于将技术规范拆解为可执行的任务清单。

  • 适合需要结构化实施计划的工程场景。engineer-plan 属于研究检索类 Skill,可作为该场景下的辅助能力补充。
  • 可自动识别技术规格和需求文档,生成有序任务。
  • 输出包含依赖关系和验收标准的详细计划。
  • 建议使用版本控制管理计划变更,确保可追溯性。

SKILL.md

/engineer-plan — Implementation Planning

Create structured implementation plans that break specs into ordered, actionable tasks. Plans bridge architecture (tech spec) and execution (engineer-work).

When to Use

  • User says "plan this", "create a plan", "implementation plan"
  • After completing a tech spec and ready to break it into tasks
  • When a feature needs structured decomposition before building

Process

Step 1: Gather Context

Check for existing specs (in priority order):

  1. docs/tech-specs/ — Most detailed, preferred source
  2. docs/prds/ — Product requirements if no tech spec
  3. User-provided description or $ARGUMENTS

Also check docs/solutions/ for relevant past solutions.

Step 2: Research (Parallel Agents)

Spawn 2 agents IN PARALLEL using the Task tool:

Agent 1 — Codebase Analyst

Task(subagent_type: "general-purpose", description: "Analyze codebase for planning")
prompt: Analyze this project to inform an implementation plan for [FEATURE].
  Focus on: existing patterns to follow, files/directories that need modification,
  related code that could be affected, test patterns, tech stack (check package.json,
  config files). Return findings relevant to planning.

Agent 2 — Practices Researcher

Task(subagent_type: "general-purpose", description: "Research best practices")
prompt: Research implementation best practices for [FEATURE TYPE] given the
  project's tech stack. Focus on: recommended implementation order, common
  integration challenges, testing strategies. Return practical guidance.

If the feature involves unfamiliar framework APIs or third-party integrations, launch an additional agent:

Agent 3 — Docs Researcher (conditional)

Task(subagent_type: "general-purpose", description: "Research framework docs")
prompt: Research [specific framework/API] documentation for [FEATURE].
  Check version compatibility with the project's package.json.
  Focus on: correct API usage, deprecation notices, common pitfalls.

Skip this agent if the feature uses well-established patterns already present in the codebase.

Step 3: Choose Detail Level

Use AskUserQuestion:

  • "What level of detail for this plan?" (header: "Detail")

- "Minimal — task list with descriptions" — For small features, <1 day - "Standard — tasks with acceptance criteria (Recommended)" — For most features, 1 day - 1 week - "Comprehensive — detailed tasks with code guidance" — For complex features, >1 week

Step 4: Write Plan

Write the plan to docs/plans/YYYY-MM-DD-<name>.md.

Planning rules:

  • Tasks ordered by dependency (what must be done first)
  • Each task is independently implementable and testable
  • Include clear acceptance criteria (standard/comprehensive)
  • Mark parallel-safe tasks
  • Effort estimates: S (single file), M (multiple files), L (cross-cutting)
  • Include Task 0 for setup (branch, dependencies) if needed

Minimal plan format:

# Plan: [Feature Name]
**Date:** YYYY-MM-DD | **Spec:** [link] | **Detail:** Minimal

## Tasks
| # | Task | Description | Effort | Depends On |
|---|------|-------------|--------|------------|
| 0 | Setup | Branch, dependencies | S | — |
| 1 | ... | ... | S/M/L | — |

Standard plan format — adds per-task sections with:

  • Key files to modify
  • Acceptance criteria (2-4 bullets)
  • Test expectations
  • Dependency graph

Comprehensive plan format — adds per-task:

  • Implementation guidance with code patterns
  • Detailed test specifications
  • Rollback procedures
  • Risk assessment

Step 5: Handoff

Use AskUserQuestion:

  • "Plan is ready. What's next?" (header: "Next step")

- "Start building (/engineer-work)" — Execute the plan - "Refine the plan" — Adjust tasks or detail - "Done for now" — Save and close

Output

Save to: docs/plans/YYYY-MM-DD-<name>.md

Next Steps

  • Ready to build? → /engineer-work
  • Need a tech spec first? → /product-tech-spec
  • Want to review after building? → /engineer-review

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

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

展示第三方安全扫描或审计结果

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

Codex

39.4%
按下载量换算25

Claude

27.61%
按下载量换算17

Cursor

17.79%
按下载量换算11

Gemini CLI

9.23%
按下载量换算6

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/elliottrjacobs/bench-skills --skill engineer-plan 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

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