Token导航 LogoToken导航

Agent Skill · 研究检索

marketing-automation(营销自动化)

marketing-automation 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。

marketing-automation 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,第三方扫描已通过,按安装量在研究检索类的 73,674 个 Skill 里排第 36,169。

总安装 1,972周安装 83GitHub Stars 11下载 691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:marketing-automation(营销自动化)
来源仓库:https://github.com/akillness/oh-my-skills
仓库路径:skills/marketing-automation
安装命令:
npx skills add https://github.com/akillness/oh-my-skills --skill marketing-automation
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/akillness/oh-my-skills --skill marketing-automation

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。

该技能用于营销自动化中的需求分类与流程引导,帮助选择正确的营销执行路径和操作模板。适用于需要明确运营模式、确定负责人、依赖关系和审批流程的营销规划场景。核心能力包括将复杂请求归一化为标准操作包,并合理路由至专业工作流,避免过度发散。安装前建议核实权限边界、维护状态及是否触发网络访问或系统操作。

marketing-automation 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/marketing-automation

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/akillness/oh-my-skills --skill marketing-autom…;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

marketing-automation 怎么安装?

在终端执行 npx skills add https://github.com/akillness/oh-my-skills --skill marketing-automation。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

marketing-automation 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。

marketing-automation 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。

marketing-automation 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。

marketing-automation 用的人多吗?

公开数据是累计安装 1,972 次,近一周安装 83 次,下载 691 次,GitHub 11 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。

marketing-automation 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

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。
  • 第三方扫描已通过,常见的高风险行为没有命中。

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
marketing-automation

营销自动化 · 本页

1,97211需要联网未明确通过

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

Marketing Automation

Use this skill as the canonical broad marketing front door when the real job is choosing the right next marketing packet.

The job is not to generate a giant campaign brainstorm. The job is to:

  1. classify the ask into one operating mode,
  2. choose one primary lane,
  3. produce one reusable operator packet,
  4. name the owner, dependencies, approvals, and proof stack,
  5. route out when the request is actually planning, game-store launch ops, or another narrower specialist workflow.

Read these support docs before finalizing the packet:

When to use this skill

  • The user asks for broad marketing help and the right next packet is still ambiguous.
  • A product, website, funnel, launch, onboarding flow, or pricing surface needs one structured marketing brief before detailed execution starts.
  • The request mixes strategy, surface choice, messaging, content, channels, and measurement.
  • The user needs a reusable operator brief, not a pile of disconnected suggestions.

When not to use this skill

  • The ask is already clearly Steam/store-page/festival/game launch work → steam-store-launch-ops.
  • The real job is backlog shaping, milestone coordination, or cross-functional implementation slicing → task-planning.
  • The user wants only one atomic copy rewrite or finished content asset with no broader routing layer → use the narrower writing/execution workflow directly.
  • The request is mostly product strategy, UX research, or technical implementation rather than marketing routing → route to the stronger adjacent skill.

Core routing model

Operating modes

Use one primary mode per run:

  • launch-orchestration — broad launch, GTM, pricing, or rollout sequencing across multiple surfaces
  • conversion-surface — landing page, pricing page, signup flow, paywall, or onboarding conversion friction
  • lifecycle-retention — onboarding, activation, retention, churn, lifecycle email, re-engagement, or user education motion
  • acquisition-content — SEO, content strategy, comparison pages, organic acquisition, or channel-facing discovery work
  • measurement-experiment — analytics setup, attribution, KPI cleanup, campaign readout, or experiment/backlog instrumentation

Primary lanes

After choosing the mode, still choose one primary lane:

  • CRO
  • Copy & messaging
  • SEO & content
  • Ads & analytics
  • Strategy & growth

The mode explains what kind of situation this is. The lane explains which marketing discipline owns the next packet.

Instructions

Step 1: Normalize the intake into one routing profile

Capture the request in this compact form before choosing the packet:

marketing_router_profile:
  primary_mode: launch-orchestration | conversion-surface | lifecycle-retention | acquisition-content | measurement-experiment
  primary_lane: CRO | Copy & messaging | SEO & content | Ads & analytics | Strategy & growth | unknown
  objective: acquisition | activation | conversion | retention | revenue | awareness | unknown
  audience:
    segment: "who this is for"
    stage: unaware | problem-aware | evaluating | active-user | churn-risk | unknown
  surface_or_channel: landing-page | pricing-page | signup-flow | onboarding | lifecycle-email | seo-page | campaign | dashboard | launch | unknown
  offer_or_motion: "product / feature / campaign / launch motion"
  measurement_maturity: clear | partial | weak
  main_question: "what needs to be decided next?"
  delivery_owner: "who will carry the packet next"
  dependencies_or_approvals:
    - "design / analytics / engineering / legal / partner / none"
  proof_assets_available:
    - "dashboard / baseline export / campaign data / user feedback / none"
  constraints:
    timeline: immediate | this-week | this-month | longer
    brand_or_compliance_notes: "limits or promises"
    domain_specificity: general | specialist | game-store

If detail is missing, proceed with explicit assumptions instead of stalling.

Step 2: Gather the minimum credible evidence

Do not route from vibes alone. Pull the smallest packet that supports a real decision:

  • objective and current bottleneck
  • audience / segment / funnel stage
  • surface, campaign, or lifecycle moment
  • available proof points or constraints
  • current KPI or best available proxy
  • who owns the next move
  • which dependencies or approvals can block execution
  • what is still missing that could change the lane choice

Step 3: Choose one mode, one lane, one packet

Use references/operating-modes-and-route-outs.md.

Default packet mapping:

  • launch-orchestration → usually Strategy & growth → launch/growth brief
  • conversion-surface → usually CRO or Copy & messaging → measurement + experiment packet
  • lifecycle-retention → usually Strategy & growth or Copy & messaging → channel-ready brief
  • acquisition-content → usually SEO & content → channel-ready brief
  • measurement-experiment → usually Ads & analytics → measurement + experiment packet

Rules:

  • Choose one primary mode.
  • Choose one primary lane.
  • Return one primary packet.
  • Mention secondary handoffs only after the main packet is chosen.

Step 4: Add route-outs before the packet sprawls

Route out instead of absorbing adjacent work when:

  • the ask is really Steam wishlists / capsules / Next Fest / store visibility → steam-store-launch-ops
  • the ask is really execution slicing / backlog organization / milestone coordination → task-planning
  • the ask is already a narrow specialist workflow with a better dedicated skill

A good front door narrows the next move. It does not win by claiming every neighboring job.

Step 5: Build the operator brief

Use references/operator-packet-and-proof-stack.md.

Return this structure:

# Marketing Routing Brief

## Intake summary
- Mode: ...
- Primary lane: ...
- Objective: ...
- Audience / stage: ...
- Surface / channel: ...
- Confidence: high | medium | low

## What matters most now
- 2-4 bullets

## Recommended packet
- Packet type: launch/growth brief | channel-ready brief | copy/messaging packet | measurement + experiment packet | marketing-routing brief
- Why this packet now: ...

## Operator packet
- Primary owner: ...
- Dependencies / approvals: ...
- Required assets or inputs: ...
- Delivery horizon: ...

## Priority decisions
| Decision | Why now | Owner | Risk if delayed |
|----------|---------|-------|-----------------|
| ... | ... | ... | ... |

## Immediate next steps
1. ...
2. ...
3. ...

## Proof stack
- Primary KPI: ...
- Leading signal: ...
- Baseline or assumption: ...
- Success threshold: ...
- Readout window: ...
- Next owner / workflow: ...

## Secondary handoffs
- Skill / workflow: ...
- Why: ...

## Not yet
- 1-3 bullets that prevent scope drift

Step 6: Keep proof attached to the packet

Every output must name:

  • one primary KPI
  • one leading signal
  • one baseline or explicit assumption
  • one success threshold
  • one readout window or review checkpoint
  • one next owner or downstream workflow

Use references/operator-packet-and-proof-stack.md for the operator packet shape and references/measurement-handoff.md for lane-specific KPI examples.

Output format

Always return a short operator-style Marketing Routing Brief.

Required qualities:

  • one primary mode
  • one primary lane
  • one packet that can actually be handed off
  • explicit assumptions when context is thin
  • named owner plus dependency/approval visibility
  • route-outs when the ask is not really this skill
  • proof logic attached to the packet, not bolted on later

Examples

Example 1: broad launch ask

Input

We are launching a new SaaS workflow next month and need help with messaging, landing page copy, onboarding emails, and how to measure whether it worked.

Output sketch

  • Mode: launch-orchestration
  • Lane: Strategy & growth
  • Packet: launch/growth brief
  • Secondary handoffs mention copy/messaging and lifecycle execution
  • Operator packet names one owner plus likely dependencies or approvals
  • Proof stack names one launch KPI and one leading signal

Example 2: pricing-page refresh

Input

Our pricing page gets traffic, but upgrades are weak. We need help with the page and what to test.

Output sketch

  • Mode: conversion-surface
  • Lane: CRO
  • Packet: measurement + experiment packet
  • Priority decisions cover offer clarity, CTA hierarchy, proof, and friction

Example 3: lifecycle / onboarding ask

Input

Trial users activate once and disappear. We need onboarding and lifecycle help, plus a sensible KPI.

Output sketch

  • Mode: lifecycle-retention
  • Lane: Strategy & growth or Copy & messaging
  • Packet: channel-ready brief
  • Operator packet names the lifecycle owner and any analytics or product dependency
  • Proof stack names activation or retained-usage KPI plus a leading signal

Example 4: game-store launch near miss

Input

We need Steam capsule copy and Next Fest marketing help for our game demo.

Output sketch

  • Keep the response short
  • Route execution to steam-store-launch-ops
  • Leave only a concise handoff note instead of a giant generic marketing plan

Best practices

  1. Act like a front-door router, not a generic growth advisor.
  2. Choose one mode before you choose tactics.
  3. Choose one packet before you list channels.
  4. Keep owner, approvals, and dependencies visible instead of assuming automation removes them.
  5. Keep launch/planning/game-store boundaries explicit.
  6. Attach proof and ownership to every packet.

References

平台分布

Codex

35.83%
按下载量换算248

Claude

28.73%
按下载量换算199

Cursor

16.8%
按下载量换算116

Gemini CLI

9.2%
按下载量换算64

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

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。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