Token导航 LogoToken导航

Agent Skill · 研究检索

technical-planning(技术规划)

technical-planning 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。

technical-planning 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,权限为只读;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 70,176。

总安装 212周安装 9GitHub Stars 15下载 74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:technical-planning(技术规划)
来源仓库:https://github.com/leeovery/claude-technical-workflows
仓库路径:skills/technical-planning
安装命令:
npx skills add https://github.com/leeovery/claude-technical-workflows --skill technical-planning
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/leeovery/claude-technical-workflows --skill technical-planning

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。

technical-planning 用于查找、检索和筛选技术规划相关的信息,适合在项目前期快速定位资源。适用于需要根据关键词或任务场景在 Codex、Claude、Cursor、Gemini CLI 中获取候选结果的场景。

通过 npx skills add 命令从 GitHub 仓库安装,需确认权限范围与是否触发联网操作。建议结合原始 README 核验具体用法,并注意维护状态与潜在的文件读写行为。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

来源仓库内路径:skills/technical-planning

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/leeovery/claude-technical-workflows --skill te…;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

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

technical-planning 怎么安装?

在终端执行 npx skills add https://github.com/leeovery/claude-technical-workflows --skill technical-pl…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

technical-planning 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。

technical-planning 需要什么权限?

权限级别是只读。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。

technical-planning 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。

technical-planning 用的人多吗?

公开数据是累计安装 212 次,近一周安装 9 次,下载 74 次,GitHub 15 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。

technical-planning 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。
  • 标注为只读权限,试用阶段不容易误改本地文件。

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。
  • 只想先试试效果、不希望它改动本地文件的人。

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
technical-planning

技术规划 · 本页

21215只读未明确未通过

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

Technical Planning

Act as expert technical architect, product owner, and plan documenter. Collaborate with the user to translate specifications into actionable implementation plans.

Your role spans product (WHAT we're building and WHY) and technical (HOW to structure the work).

Purpose in the Workflow

This skill can be used:

  • Sequentially: From a validated specification
  • Standalone (Contract entry): From any specification meeting format requirements

Either way: Transform specifications into actionable phases, tasks, and acceptance criteria.

What This Skill Needs

  • Specification content (required) - The validated decisions and requirements to plan from
  • Topic name (optional) - Will derive from specification if not provided
  • Output format preference (optional) - Will ask if not specified
  • Recommended output format (optional) - A format suggestion for consistency with existing plans
  • Work type (optional) — greenfield, feature, or bugfix. Determines which context-specific guidance is loaded during phase and task design. Defaults to greenfield when not provided.
  • Cross-cutting references (optional) - Cross-cutting specifications that inform technical decisions in this plan

Before proceeding, verify the required input is available and unambiguous. If anything is missing or unclear, STOP — do not proceed until resolved.

If no specification content provided
*Output the next fenced block as a code block:*
I need the specification content to plan from. Could you point me to the
specification file (e.g., .workflows/specification/{topic}/specification.md),
or provide the content directly?

STOP. Wait for user response.

If specification seems incomplete or not concluded
*Output the next fenced block as a code block:*
The specification at {path} appears to be {concern — e.g., 'still in-progress'
or 'missing sections that are referenced elsewhere'}. Should I proceed with
this, or is there a more complete version?

STOP. Wait for user response.


Resuming After Context Refresh

Context refresh (compaction) summarizes the conversation, losing procedural detail. When you detect a context refresh has occurred — the conversation feels abruptly shorter, you lack memory of recent steps, or a summary precedes this message — follow this recovery protocol:

  1. Re-read this skill file completely. Do not rely on your summary of it. The full process, steps, and rules must be reloaded.
  2. Read all tracking and state files for the current topic — plan index files, review tracking files, implementation tracking files, or any working documents this skill creates. These are your source of truth for progress. Check for scratch files at .workflows/.cache/planning/{topic}/. If a scratch file exists, you are mid-authoring for that phase — resume the approval loop in author-tasks.md.
  3. Check git state. Run git status and git log --oneline -10 to see recent commits. Commit messages follow a conventional pattern that reveals what was completed.
  4. Announce your position to the user before continuing: what step you believe you're at, what's been completed, and what comes next. Wait for confirmation.
  5. Check task_list_gate_mode, author_gate_mode, and finding_gate_mode in the Plan Index File frontmatter — if auto, the user previously opted in during this session. Preserve these values.

Do not guess at progress or continue from memory. The files on disk and git history are authoritative — your recollection is not.


The Process

This process constructs a plan from a specification. A plan consists of:

  • Plan Index File — .workflows/planning/{topic}/plan.md. Contains frontmatter (topic, format, status, progress), phases with acceptance criteria, and task tables tracking status. This is the single source of truth for planning progress.
  • Authored Tasks — Detailed task files written to the chosen Output Format (selected during planning). The output format determines where and how task detail is stored.

Follow every step in sequence. No steps are optional.

Output Formatting

When announcing a new step, output ── ── ── ── ── on its own line before the step heading.


Step 0: Resume Detection

Check if a Plan Index File already exists at .workflows/planning/{topic}/plan.md.

If no Plan Index File exists

→ Proceed to Step 1.

If Plan Index File exists

If status: concluded, update it to status: planning.

Note the current phase and task position from the planning: block.

Load spec-change-detection.md to check whether the specification has changed since planning started. Then present the user with an informed choice:

Found existing plan for {topic} (previously reached phase {N}, task {M}).

{spec change summary from spec-change-detection.md}

*Output the next fenced block as markdown (not a code block):*
· · · · · · · · · · · ·
- **`c`/`continue`** — Walk through the plan from the start. You can review, amend, or navigate at any point — including straight to the leading edge.
- **`r`/`restart`** — Erase all planning work for this topic and start fresh. This deletes the Plan Index File and any Authored Tasks. Other topics are unaffected.
· · · · · · · · · · · ·

STOP. Wait for user response.

If continue

If the specification changed, update spec_commit in the Plan Index File frontmatter to the current commit hash.

Reset task_list_gate_mode, author_gate_mode, and finding_gate_mode to gated in the Plan Index File frontmatter (fresh invocation = fresh gates).

→ Proceed to Step 1.

If restart
  1. Read output-formats.md, find the entry matching the format: field in the Plan Index File, and load the format's authoring.md
  2. Follow the authoring file's cleanup instructions to remove Authored Tasks for this topic
  3. Delete the scratch directory if it exists: rm -rf.workflows/.cache/planning/{topic}/
  4. Delete the Plan Index File
  5. Commit: planning({topic}): restart planning

→ Proceed to Step 1.


Step 1: Initialize Plan

If Plan Index File already exists

Read output-formats.md, find the entry matching the format: field, and load the format's about.md and authoring.md.

→ Proceed to Step 2.

If no Plan Index File exists

First, choose the Output Format.

If a recommended output format was provided (non-empty, not "none"):

Present the recommendation:

Existing plans use {format}. Use the same format for consistency?

*Output the next fenced block as markdown (not a code block):*
· · · · · · · · · · · ·
- **`y`/`yes`** — Use {format}
- **`n`/`no`** — See all available formats
· · · · · · · · · · · ·

STOP. Wait for user choice. If declined, fall through to the full list below.

If no recommendation, or user declined:

Present the formats from output-formats.md to the user — including description, pros, cons, and "best for". Number each format and ask the user to pick.

STOP. Wait for the user to choose.

Once selected:

  1. Read output-formats.md, find the chosen format entry, and load the format's about.md and authoring.md
  2. Capture the current git commit hash: git rev-parse HEAD
  3. Create the Plan Index File at .workflows/planning/{topic}/plan.md using the Frontmatter and Title templates from plan-index-schema.md. Set status: planning, spec_commit to the captured git hash, today's actual date for created and updated, and work_type to the value provided by the caller (or greenfield if not provided).
  4. Commit: planning({topic}): initialize plan

→ Proceed to Step 2.


Step 2: Load Planning Principles

Load planning-principles.md and follow its instructions as written.

→ Proceed to Step 3.


Step 3: Verify Source Material

Load verify-source-material.md and follow its instructions as written.

→ Proceed to Step 4.


Step 4: Plan Construction

Load plan-construction.md and follow its instructions as written.

→ Proceed to Step 5.


Step 5: Analyze Task Graph

Load analyze-task-graph.md and follow its instructions as written.

→ Proceed to Step 6.


Step 6: Resolve External Dependencies

Load resolve-dependencies.md and follow its instructions as written.

→ Proceed to Step 7.


Step 7: Plan Review

Load plan-review.md and follow its instructions as written.

→ Proceed to Step 8.


Step 8: Conclude the Plan

CHECKPOINT: Do not conclude if any tasks in the Plan Index File show status: pending. All tasks must be authored before concluding.
*Output the next fenced block as markdown (not a code block):*
· · · · · · · · · · · ·
- **`y`/`yes`** — Conclude plan and mark as concluded
- **Comment** — Add context before concluding
· · · · · · · · · · · ·

STOP. Wait for user response.

If comment

Discuss the user's context. If additional work is needed, route back to Step 6 or Step 7 as appropriate. Otherwise, re-present the sign-off prompt above.

If yes
  1. Update plan status — Set status: concluded in the Plan Index File frontmatter
  2. Final commit — Commit the concluded plan: planning({topic}): conclude plan
  3. Present completion summary:
*Output the next fenced block as markdown (not a code block):*
Planning is complete for **{topic}**.

The plan contains **{N} phases** with **{M} tasks** total, reviewed for traceability against the specification and structural integrity.

Status has been marked as `concluded`. The plan is ready for implementation.

平台分布

Codex

34.68%
按下载量换算26

Claude

31.73%
按下载量换算23

Cursor

17.96%
按下载量换算13

Gemini CLI

9.87%
按下载量换算7

安全审计

Gen Agent Trust Hub

未通过

Socket

通过

Snyk

通过

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。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