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审计提醒

marketing-social-media营销社交媒体

Agent Skill

marketing-social-media 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

1,608

周安装

67

GitHub Stars

59

下载量

536
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

3

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:marketing-social-media(营销社交媒体)
来源仓库:https://github.com/vasilyu1983/ai-agents-public
仓库路径:skills/marketing-social-media
安装命令:
npx skills add https://github.com/vasilyu1983/ai-agents-public --skill marketing-social-media
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/vasilyu1983/ai-agents-public --skill marketing-social-media

简介

用于查找、检索和筛选社交媒体平台相关信息。

  • 适合在 Codex、Claude、Cursor、Gemini CLI 中快速定位候选结果。
  • 通过关键词或任务场景进行信息检索与筛选。marketing-social-media 属于研究检索类 Skill,可作为该场景下的辅助能力补充。
  • 安装前需确认权限范围和维护状态,避免触发联网或文件读写。
  • 建议结合原始 README 核验具体用法和功能边界。

SKILL.md

Social Media Marketing (Operational Skill)

This skill contains actionable, non-theoretical operational systems for sustainable social media marketing across LinkedIn, TikTok, Meta (FB/IG), YouTube, and X.

Structure: Core social media principles first. AI-specific automation in clearly labeled "Optional: AI / Automation" sections.

Scope note: For email outreach and lead generation, see marketing-leads-generation. For SEO, see marketing-seo-complete.

How to use this skill

  • Default workflow: pick goal -> pick 1-2 platforms -> set a sustainable cadence -> publish -> engage daily -> review weekly -> iterate monthly.
  • Use assets/ for ready-to-run templates; use references/ for deep dives (platform, paid, community, AI governance).
  • For time-sensitive platform changes, use the Trend Awareness Protocol.

Core: Platform-Agnostic Principles

These principles apply regardless of platform. Master them before platform-specific tactics.

The 3 Pillars

PillarDescriptionMeasure
AudienceWho you serve, their problems, what they valueFollower quality, engagement rate
FormatContent type optimized for platform + audienceView-through, completion, saves
CadenceConsistent publishing rhythmPosting frequency, time-to-response

Audience-First Thinking

QuestionWhy It Matters
Who specifically am I creating for?Generic content = generic results
What problem does this solve for them?Value-first, not self-promotion
Where do they spend time online?Platform selection follows audience
What format do they prefer?Match format to consumption habits

Format Selection by Goal

GoalBest FormatsWorst Formats
AwarenessShort-form video, carousels, threadsLong-form without distribution
EngagementPolls, questions, controversial takesCorporate announcements
AuthorityLong-form, original research, case studiesRepurposed generic content
ConversionDirect CTAs, testimonials, demosSoft branding

Cadence Guidelines

Start with a cadence you can sustain for 8-12 weeks without sacrificing quality.

Cadence TierTypical OutputBest ForNotes
Light2-3 posts/weekSmall teams, early consistencyUse templates + repurposing
Standard4-7 posts/weekMost product teamsMix educational + proof + community
HeavyDaily + community touchpointsMedia-led teamsRequires strong production pipeline

Core: Content Typology

Every post should map to one of these types. Mix types across your content calendar.

TypePurposeExampleFrequency
EducationalTeach something valuableHow-to, tutorial, framework40-50%
NarrativeShare stories that resonatePersonal journey, case study, behind-the-scenes20-30%
ProofDemonstrate results/credibilityTestimonial, metric, achievement15-20%
CommunityEngage and connectQuestion, poll, reply, shoutout10-15%
PromotionalDrive actionOffer, CTA, launch5-10% max

Do (Content Mix)

  • Plan content mix monthly, not post-by-post
  • Lead with educational; end week with proof or community
  • Test different type ratios for your audience
  • Repurpose across types (educational -> narrative)

Avoid (Content Mix)

  • Mostly promotional content (looks like spam)
  • All educational (no personality, no engagement)
  • Random posting without type intention
  • Copying competitor content 1:1

Core: Engagement Quality vs Vanity Metrics

What to Track (Quality)

MetricWhy It Matters
Saves/bookmarksStrong signal that content is useful
Shares/repostsIndicates distribution value and message resonance
Comment qualityDepth of discussion and objections surfaced
Profile visitsInterest in the creator/company after content
Link clicksConversion path signal (only if UTMs are clean)
Downstream conversionsSign-ups, demos, trials, purchases from social landings

What to Ignore (Vanity)

MetricWhy It's Misleading
Follower count aloneBots, inactive, wrong audience
Impressions without engagementSeen!= valued
Likes without contextEasy to give, low signal
Viral posts without conversionFame!= revenue

Do (Metrics)

  • Track engagement rate, not raw likes
  • Monitor save rate as quality signal
  • Measure conversion from social (UTMs, ask "how did you hear about us")
  • Compare metrics to your own baseline, not others

Avoid (Metrics)

  • Buying followers or engagement
  • Chasing viral at expense of relevance
  • Ignoring negative engagement signals (unfollows, hides)
  • Reporting vanity metrics to stakeholders

Core: Revenue Attribution & ROI

Transform social from cost center to revenue driver by measuring business impact.

Metric CategoryKey MetricsSee Reference
Social-to-RevenueSocial CAC, LTV:CAC, pipeline influencerevenue-attribution-framework.md
Influencer ROICPA, ROAS, EMV, content asset valuerevenue-attribution-framework.md
Community ROIMember LTV premium, support deflection, retention deltarevenue-attribution-framework.md
Brand LiftUnaided/aided awareness, consideration, preferencerevenue-attribution-framework.md

Quick Implementation

  1. UTM discipline: utm_source, utm_medium, utm_campaign, utm_content, utm_term on all links
  2. CRM integration: Pass UTM data on conversion, tag closed deals with source
  3. Cohort analysis: Compare LTV by acquisition source
  4. Attribution model: Choose first-touch (awareness) or position-based (balanced)

Key Benchmarks

MetricTarget
Social LTV:CAC>3:1
Influencer ROAS>2x (direct), >4x (with EMV)
Community member LTV1.2-2x non-member
Support deflection20-40%

For detailed frameworks, scorecards, and reporting templates, see references/revenue-attribution-framework.md.

Core: Social -> Site -> Conversion Flows

Social media is rarely the conversion point. Design the flow:

Social Post -> Link/Bio -> Landing Page -> Conversion

OR

Social Post -> DM/Comment -> Conversation -> Conversion

OR

Social Post -> Lead Magnet -> Email Nurture -> Conversion

Flow Design Checklist

StepQuestionIf No
PostDoes it provide standalone value?Rewrite - no one clicks without value
CTAIs there one clear next action?Simplify to single CTA
Link/BioDoes it match post promise?Align messaging
LandingDoes it load fast, mobile-first?Optimize (see marketing-seo-complete)
ConversionIs friction minimized?Reduce form fields, add trust signals

Use content-distribution-plan.md to plan search + social distribution for each piece of content.

Do (Conversion Flows)

  • Test bio link vs in-post link
  • Use UTM parameters to track social -> site
  • Optimize landing page for mobile-first reading and fast load
  • Retarget engaged users who didn't convert

Avoid (Conversion Flows)

  • Generic bio links that don't change
  • No tracking on social traffic
  • Desktop-first landing pages
  • Sending cold traffic straight to purchase

Core: Community Management Basics

Response Time Standards

Interaction TypeTarget Response Time
Comment on your postSame day (faster during launch windows)
Direct messageWithin 1 business day (faster for high-intent inquiries)
Mention/tagWithin 24 hours
Negative feedbackASAP (same day)

Moderation Principles

SituationAction
Spam/botDelete and block
Constructive criticismThank, address publicly
Negative but validAcknowledge, offer to continue privately
Trolling/harassmentHide/delete, block if repeated
QuestionsAnswer helpfully, link to resources

Do (Community)

  • Respond to every real comment (first week especially)
  • Ask follow-up questions to drive discussion
  • Highlight community members (reshare, mention)
  • Set community guidelines for owned spaces

Avoid (Community)

  • Ignoring comments (signals you don't care)
  • Deleting all criticism (looks like hiding)
  • Generic copy-paste responses
  • Letting trolls derail conversations

Platform Fit Notes (2026)

PlatformBest ForPrimary FormatsKey notes
LinkedInB2B, authorityText posts, carousels, native videoDwell time and meaningful comments matter; employee amplification can materially help
XNews, opinions, real-timeShort posts, threads, quotesFast half-life; write for context collapse (assume zero prior context)
TikTokDiscovery, social commerceVertical video, live, TikTok ShopRetention is the core signal; test multiple lengths and hooks
InstagramVisual storytelling, creatorsReels, carousels, storiesReels/carousels drive reach; make the first 2 seconds and cover/frame clear
YouTubeEvergreen + searchLong-form + Shorts + ShoppingRetention + session time matter; titles/thumbnails are part of the product
RedditHigh-trust communitiesText posts, AMAs, participationAuthentic voice required; follow subreddit norms; avoid drive-by promotion
ThreadsConversation (Meta ecosystem)Short posts, threadsStrong tie-in with IG; treat as community/engagement channel, not a link farm
PinterestDiscovery, shopping intentPins, idea pinsStrong for evergreen product/content and visual search categories

For detailed platform-specific guidance, see:

Use this skill when the user asks for:

  • content ideas, calendars, scripts, hooks
  • ad creative templates
  • growth frameworks
  • optimization workflows
  • paid social setup and scaling
  • copywriting requests
  • community management
  • performance audits
  • revenue attribution and ROI measurement
  • influencer program ROI
  • community value calculation
  • brand lift studies

Decision Trees

Paid Social: Low CTR -> Refresh hook/creative | Clicks ok, CVR low -> Fix landing page | Costs rising -> Refresh creative
Organic: Inconsistent -> Set sustainable cadence | Low engagement -> Improve value density | Low clicks -> Strengthen CTA
Community: Spam rising -> Enforce guidelines | Low retention -> Add rituals | Negative sentiment -> Acknowledge publicly

Navigation

References (Deep Dives)

Templates

CategoryLocation
Adsassets/ads/ - Google RSA, Meta, TikTok, YouTube
Emailassets/email/ - Cold, nurture, newsletter, sales
Socialassets/social/ - Calendars, hooks, carousels, scripts
Offersassets/offers/ - Offer builder, value stack, guarantees
Communityassets/community/ - Strategy, guidelines, events, moderation
SEOassets/seo/ - Briefs, outlines, keyword research

Related Skills

External Resources

See data/sources.json for official documentation across Google Ads, Meta, TikTok, YouTube, and platform analytics (automation sources are tagged optional).

Safety and compliance (required)

  • Follow each platform's policies and ad standards; verify disclosure requirements for sponsored/branded content in your market.
  • Do not recommend buying followers/engagement, deceptive tactics, or automation that impersonates humans.
  • Avoid storing or sharing personal data; use aggregated metrics and consent-based tracking.

Optional: AI / Automation

Note: Core social media fundamentals above work without AI tools. This section covers optional automation capabilities.

Scheduling & Publishing

ToolPlatformsBest For
BufferAll majorSimple scheduling, analytics
HootsuiteAll majorEnterprise, team collaboration
LaterIG, TikTok, PinterestVisual planning, link-in-bio
Sprout SocialAll majorEnterprise, CRM integration
Native schedulersPlatform-specificBest reach (some algorithms favor native)

AI Content Assistance

Use CaseToolsConsideration
IdeationChatGPT, ClaudeStarting point only - add human insight
Hook generationLLM promptsA/B test AI vs human hooks
RepurposingOpus Clip, DescriptVideo -> clips, long -> short
Image generationMidjourney, DALL-ECheck platform policies on AI content
Caption writingLLM promptsAlways edit for brand voice

Moderation & Monitoring

Use CaseToolsBenefit
Comment moderationMeta AI, SprinklrFilter spam, flag issues
Brand monitoringMention, Brand24Track mentions across platforms
Sentiment analysisSprout, BrandwatchAggregate sentiment trends

Do (AI/Automation)

  • Use scheduling to maintain consistency when you can't post live
  • Test AI-generated hooks/captions against human-written
  • Automate monitoring, not engagement (replies should be human)
  • Label AI-generated content where required by platform

Avoid (AI/Automation)

  • Full automation of replies/DMs (obvious, damages trust)
  • AI content without human review (brand voice, factual accuracy)
  • Over-relying on AI ideation (becomes generic)
  • Ignoring platform policies on AI content disclosure

See references/ai-marketing-governance.md for AI governance framework.

Collaboration Notes

With Product

  • Launch coordination: Social amplification for product launches
  • Feature education: Tutorial content for new features
  • User feedback: Social comments/DMs = product insights
  • Community features: Product roadmap for community needs

With Sales

  • Social selling support: Content for sales team to share
  • Lead alerts: DMs and comments with buying signals -> Sales
  • Case studies: Sales wins -> Social proof content
  • Event promotion: Webinars, demos, conferences

With Engineering

  • Social embeds: Share buttons, social proof widgets
  • API integrations: CRM sync, lead capture from social
  • Analytics setup: UTM handling, conversion tracking
  • Page performance: Social traffic = mobile traffic = fast pages

With Design

  • Visual templates: Consistent brand assets across platforms
  • Video production: Reels, TikToks, YouTube Shorts
  • Carousel design: LinkedIn, Instagram carousel templates
  • UGC guidelines: Brand guidelines for community content

Trend Awareness Protocol

IMPORTANT: When users ask time-sensitive recommendation questions about social media marketing, verify current platform changes before answering.

  • If a WebSearch tool is available, use it.
  • Prefer official docs and policy pages listed in data/sources.json, then validate with up-to-date reporting.
  • If you cannot use WebSearch, state that you are using stable principles and propose an experiment plan to validate assumptions.

Trigger Conditions

  • "What's the best social media strategy for 2026?"
  • "What should I focus on for [LinkedIn/TikTok/Instagram]?"
  • "What's trending on [platform]?"
  • "Current best practices for [Reels/Shorts/Stories]?"
  • "Is [platform/format] still relevant in 2026?"
  • "How has the [LinkedIn/TikTok/Instagram] algorithm changed?"
  • "What content format works best on [platform]?"

Required Searches

  1. Search: "social media marketing trends 2026"
  2. Search: "[LinkedIn/TikTok/Instagram] algorithm 2026"
  3. Search: "[platform] best practices January 2026"
  4. Search: "social media engagement trends 2026"

What to Report

After searching, provide:

  • Current landscape: What platforms/formats work NOW
  • Emerging trends: New features or platforms gaining traction
  • Deprecated/declining: Strategies or formats losing effectiveness
  • Recommendation: Based on fresh data, not just static knowledge

Example Topics (verify with fresh search)

  • Platform algorithm changes (LinkedIn, TikTok, Instagram, X)
  • New content formats (Threads, Notes, Shorts features)
  • Creator economy trends and monetization
  • AI content creation tools for social
  • Community features and engagement patterns
  • Cross-platform posting best practices

International Markets

This skill covers global platforms (Meta, LinkedIn, TikTok). For regional platforms:

NeedSee Skill
China platforms (WeChat, Douyin, Weibo, RED)marketing-geo-localization
Japan/Korea (LINE, Naver, Kakao)marketing-geo-localization
Russia/CIS (VK, Telegram, Yandex)marketing-geo-localization
Cultural content adaptationmarketing-geo-localization

If your query mentions regional platforms or specific countries, also use marketing-geo-localization.

Anti-Patterns

Anti-PatternWhy It FailsInstead
Post and ghostNo feedback loop; weak trustSet a response SLA and seed discussion
Same content everywhereIgnores platform cultureAdapt format and tone per platform
Chasing followersVanity metric, often wrong audienceFocus on engagement rate, conversions
Inconsistent postingBreaks continuity and learningsBatch content, reuse templates, schedule
All promotionalAudience tunes outMake promotion the minority; lead with value
Ignoring analyticsCan't improve what you don't measureWeekly metric review, monthly deep dive
Copying competitorsNo differentiation, always behindStudy competitors, create unique angle
Buying engagementPlatform detection, fake audienceBuild organic, slower but real
No conversion pathSocial activity without business resultsClear CTA, tracked links, landing pages
Platform dependencyDistribution riskBuild owned channels (email, community) and retargeting

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

能力 5

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

Claude Code

29.54%
按下载量换算158

Antigravity

21.13%
按下载量换算113

Gemini CLI

18.82%
按下载量换算101

OpenCode

11.48%
按下载量换算62

Cursor

7.44%
按下载量换算40

windsurf

3.33%
按下载量换算18

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

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 warning/failed 结果,不能写成本站确认安全。

来源信息

继续浏览同类 Skills