Content & Media Creation · coreyhaines31/marketingskills
marketing-skills
When the user wants to create or update their product marketing context document. Also use when the user mentions 'product context,' 'marketing context,' 'set up context,' 'positioning,' 'who is my target audience,' 'describe my product,' 'ICP,' 'ideal customer profile,' or wants to avoid repeating foundational information across marketing tasks. Use this at the start of any new project before using other marketing skills — it creates `.agents/product-marketing.md` that all other skills reference for product, audience, and positioning context.
L4 High-Privilege · Graded auth & full auditV1 Commit-PinnedExecute CommandNetwork AccessWrite FileDual-pass Audit
Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.
1How It Works & Mechanism
It is a collection of 50 standalone skills (skills/<name>/SKILL.md) connected via shared context: product-marketing acts as the foundation, and all other skills read the product/audience/positioning document it generates before proceeding with their tasks.
README.md
Skills reference each other and build on shared context. The `product-marketing` skill is the foundation — every other skill checks it first to understand your product, audience, and positioning before doing anything.
Materialized as a single file at .agents/product-marketing.md (Workflow Step 1 of product-marketing checks if it exists and provides an option to auto-draft V1 from the codebase).
Every skill is written in pure Markdown instructions (no executables under skills/), implemented through domain-expert checklists and structured output disciplines; knowledge-intensive sections are modularized into references/ and loaded on demand.
skills/cro/SKILL.md
You are a conversion rate optimization expert. Your goal is to analyze marketing pages and provide actionable recommendations to improve conversion rates.
The 50 SKILL.md files total 15,196 lines, with none exceeding 500 lines based on testing (conforming to the self-imposed limit in AGENTS.md); an additional 178 references/ files hold detailed context.
The actual execution layer lives in tools/clis: 64 zero-dependency, single-file Node CLIs. Each reads credentials from environment variables, calls the vendor's API using fetch, and provides a built-in --dry-run flag to echo the request without sending.
tools/clis/ga4.js
if (args['dry-run']) { return { _dry_run: true, method, url: `${baseUrl}${path}`, headers: { Authorization: '***', 'Content-Type': 'application/json' }, body: body || undefined } }
Verified across three metrics: grep -l matches 64 files for 'process.env', 'await fetch', and 'dry-run' alike (100% coverage). If credentials are missing, the CLI exits immediately with an error (e.g., 'GA4_ACCESS_TOKEN environment variable required').
Tool knowledge is split into two layers: a registry and per-tool integration guides. REGISTRY.md provides the index and capability matrix, while tools/integrations/*.md details endpoints, authentication, and common operations.
tools/REGISTRY.md
Quick reference for AI agents to discover tool capabilities and integration methods.
Verified 95 guides under tools/integrations/; the table in REGISTRY marks API/MCP/CLI/SDK support status for each tool.
Each skill includes its own evaluation test cases (evals/evals.json), hardcoding expected outputs and assertable criteria—a rare self-testing asset for content-focused skills.
skills/cro/evals/evals.json
Checks for product-marketing.md
One file per skill across all 50 skills (find -name evals.json returns 50); test cases include prompt, expected_output, and assertions arrays (e.g., requiring output to contain Quick Wins / High-Impact Changes / Test Ideas sections).
Distribution and manifest consistency are automated: CI validates SKILL.md changes against a modification matrix on updates, and syncs marketplace.json / plugin.json / README upon push to main.
.github/workflows/validate-skill.yml
uses: Flash-Brew-Digital/validate-skill@v1
sync-skills.yml automatically commits as Coreybot with 'chore: sync skills with marketplace.json, plugin.json, and README'; local environments also provide validate-skills.sh (custom specification checks) and validate-skills-official.sh (pulls and verifies against the official skills-ref repo).
Versioning discipline is two-tiered and drives user-side update checks via VERSIONS.md: repo release versions map to plugin.json/marketplace.json/VERSIONS.md, while each skill maintains its own metadata.version.
AGENTS.md
**Per-skill version** — `metadata.version` in each SKILL.md, mirrored in the `VERSIONS.md` table. Bump on ANY shipped change to that skill: the update check compares `VERSIONS.md` against users' local skill metadata, so an unbumped change is invisible to installed users.
Update check procedure is defined in AGENTS.md: 'Fetch `VERSIONS.md` from GitHub: https://raw.githubusercontent.com/coreyhaines31/marketingskills/main/VERSIONS.md', prompting a non-blocking notice only when ≥2 skills have updates or any major version bump occurs.
Sponsorships and tool recommendations are explicitly decoupled and codified into governance documentation: partners represent an 'already disclosed presence' rather than paid recommendation weighting.
tools/PARTNERS.md
**Sponsorship funds the work, never the recommendations.** Money buys a *disclosed presence*, never a bias.
Both the Partners section in README and the Verified Partners table in REGISTRY are generated from partners.json via scripts/sync-partners.mjs (marked with <!-- PARTNERS:START/END -->), featuring a --check mode for CI verification.
2Core Capabilities
01CRO and landing page conversion diagnostics (categorized by page type, covering value propositions/headlines/CTA hierarchy/trust signals/objection handling, outputting Quick Wins / High-Impact / Test Ideas)
02Creation and maintenance of a shared product marketing context document (including versioning and changelog, preventing redundant foundational prompts across skills)
03Copywriting and content generation: copywriting / copy-editing / cold-email (B2B cold emails + follow-up sequences + subject lines with benchmark reference data)
04SEO suite: seo-audit, ai-seo, programmatic-seo, schema, site-architecture (including international SEO and AI writing detection guidelines)
05Paid advertising and experimentation: ads / ad-creative / ab-testing (including sample size guidelines and testing templates)
06Growth and retention: referrals / free-tools / churn-prevention / paywalls / popups / onboarding / signup
07GTM and revenue operations: pricing / launch / revops / sales-enablement (including objection libraries, deck frameworks, demo scripts) / prospecting
0864 marketing platform CLIs: GA4 / Mixpanel / Segment / Amplitude / Adobe Analytics / Plausible to Meta Ads / Google Ads / LinkedIn Ads / TikTok Ads, Klaviyo / Customer.io / Resend / Mailchimp, Ahrefs / Semrush / DataForSEO / G2, etc.
09Strategy and communications: AARRR framework marketing plans, PR and earned media, offer design, event marketing and webinars, co-marketing, competitive profiling
3External Dependencies
| Kind | Dependency / Endpoint |
|---|---|
| cli | node（64 个零依赖 CLI 的运行时，要求 Node 18+） |
| api | GA4（Data API + Admin API + Measurement Protocol；凭证 GA4_ACCESS_TOKEN） |
| api | Resend（邮件发送；凭证 RESEND_API_KEY） |
| api | Dub（短链/归因；凭证 DUB_API_KEY） |
| api | 其余厂商 API（每个 CLI 一个固定 base URL，共 64 个；鉴权与操作见 tools/integrations/*.md） |
| network | 更新检查（skill 指示 agent 抓取 raw.githubusercontent 上的 VERSIONS.md 比对版本） |
| network | 本地官方校验脚本会克隆并安装官方 skills-ref 库（仅维护者手动运行） |
| network | Composio（为无原生 MCP 的工具提供 MCP 访问的集成层，纯文档指引） |
| package | CI 侧第三方 Action：Flash-Brew-Digital/validate-skill@v1、actions/checkout@v6、stefanzweifel/git-auto-commit-action@v7 |
4Recommended Governance & Evidence
Suggested Controls for Safe Operation
For Agent Runtime
Graded/per-use authorization, sandbox recommended
For Administrators
Approval workflow, rollback support, full audit
Evidence & Triggers (File Paths & Excerpts)
- scripts/: Local script or command execution: tools/clis/*.js (64 files): invoked on-demand by the agent or user via `node tools/clis/<name>.js <cmd> ...`
- 网络端点: Declared external network access: 64 CLIs under tools/clis send HTTPS requests to vendor APIs (GA4, Resend, Dub, Meta/Google/LinkedIn, etc.)
- 本地工作目录: File I/O and artifact output: Core skills instruct writing to user project files, notably .agents/product-marketing.md (the primary output of product-marketing, containing versioning and context)
Audit Notes: Rating justification (based on the 6-tier classification: Orange if credentials/environment variables are accessed): 1) All 64 scripts under tools/clis read vendor API credentials from environment variables and transmit them outward via Authorization headers (e.g., GA4_ACCESS_TOKEN, RESEND_API_KEY, META_ACCESS_TOKEN, GOOGLE_ADS_TOKEN), representing real credential exposure and external data egress; 2) The skill instructs the agent to fetch VERSIONS.md from raw.githubusercontent for update checks; 3) The distribution pipeline relies on third-party GitHub Actions (Flash-Brew-Digital/validate-skill@v1, stefanzweifel/git-auto-commit-action@v7), and the local official validation script executes git clone and pip install on external repositories. Mitigations (do not lower the tier, but must be explicitly documented): The 50 core skills themselves contain zero executable scripts and read no credentials; CLI credentials originate strictly from the user's local environment, target vendor official endpoints, support --dry-run simulation, and are never written to disk; the repository does not access keychain/.env/cookies; sponsorship recommendations have strict governance and disclosure (PARTNERS.md + partners.json + generation scripts + CI checks). Cautionary guidelines: 1) Letting agents invoke these CLIs exposes third-party API keys to the session process environment—ensure proper scoping and credential rotation; 2) Always use --dry-run to inspect requests before live execution, especially for CLIs that trigger external side effects (e.g., resend, meta-ads, google-ads); 3) When using AGENTS.md's `!command` syntax in Claude Code, note that this executes commands within your own local project environment and must not be written into shared SKILL.md files.
5Second-Pass Verification
- [ok] skill.path 定位（任务表标『待定位』）；合集条目指向哪个具体 skill — 本条目展示名『Marketing Skills』对应的是 50 个 skill 的合集，仓库根无集合级 SKILL.md。skill.path 现指向 skills/product-marketing —— README 明写它是地基（'The `product-marketing` skill is the foundation — every other skill checks it first'），它产出全部其它 skill 共用的 .agents/product-marketing.md，因此最能代表这个集合的整体形态。其余 49 个 skill 并列于 skills/<name>/（cro、copywriting、seo-audit、ads、analytics、referrals、revops 等）；本 JSON 的 capabilities/traits 按合集口径书写，evidence 分别指向 skills/ 下与 tools/ 下的具体文件。
- [discrepancy] official_desc 取哪一份（合集无集合级 frontmatter description） — 合集在仓库根没有集合级 SKILL.md/frontmatter description（50 个 skill 各有自己的 description 与 metadata.version）。本条目 official_desc 采用代表 skill skills/product-marketing 的 frontmatter description 逐字原文（与 skill.path 指向的具体 skill 一致，满足『official_desc = SKILL.md frontmatter 原文』的取证规则）；集合级口径（plugin.json 的 'Marketing skills for AI agents — conversion optimization, copywriting, SEO, paid ads, ad creative, and growth' 与 marketplace.json 的 50-skill 长描述）已逐字记入 internal_assets 的清单条目说明，不丢失信息。
- [ok] 『核心 skill 无脚本』结论的反例搜索 — find skills -name '*.py' -o -name '*.sh' -o -name '*.js' 零命中；skills/ 下可执行资产为 0，全部内容为 .md 与 evals.json。所有可执行代码集中在 tools/clis（64）、scripts/（sync-partners.mjs）与两个根级 validate-skills*.sh。
- [ok] 『每个 CLI 都读凭证 + 都有 dry-run』是否全量成立 — grep -l 'process.env' tools/clis/*.js = 64；grep -l 'await fetch' = 64；grep -l 'dry-run' = 64。抽样读 ga4.js（三端点 + dry-run 打码 Authorization）、resend.js（api.resend.com）、dub.js（api.dub.co）确认模式一致，非个别实现。
- [ok] 赞助是否会污染推荐（治理是否只是口号） — PARTNERS.md 首条原则逐字 'Sponsorship funds the work, never the recommendations.'；README 与 REGISTRY 的伙伴区块由 partners.json 经 sync-partners.mjs 生成且带 ◆ 披露标记；REGISTRY 明写伙伴『listed alongside the neutral options for the same job, never instead of them』。属可核查的机制而非纯声明（生成脚本 + 标记区块 + --check）。
- [ok] 更新检查的真实动作 — AGENTS.md 'Checking for Updates' 小节逐字给出 raw.githubusercontent 上的 VERSIONS.md 抓取，并要求『Once per session, on first skill use』且仅在 ≥2 skill 有更新或主版本跳变时以非阻塞方式提示；属指示 agent 去抓取，而非仓库自带定时任务。
- [ok] CI 与本地校验的真实命令 — validate-skill.yml 用 matrix + Flash-Brew-Digital/validate-skill@v1 按改动的 SKILL.md 逐个校验；sync-skills.yml 跑 node .github/scripts/sync-skills.js 并自动提交三处清单；validate-skills.sh 为自研 shell 检查（exit 1 on issues）；validate-skills-official.sh 会 git clone github.com/agentskills/agentskills 到 /tmp 并 uv sync / pip install -e，属维护者手动运行的网络+远程包路径。
- [ok] stars / last_push 两源比对 — gh api 实时 stars=50403、pushed_at=2026-09-05T04:48:10Z；queue-100.json 快照为 50403 / 2026-09-05，一致。
6Verdict & Scenarios
Best for: Best suited for: Technical marketers and indie founders who want AI agents to directly query metrics (GA4/Ads/SEO) and deliver copy, landing pages, cold emails, and CRO audits; as well as skill authors looking to replicate an engineering pattern of breaking domain expertise into 50 cross-referencing skills, executable CLIs, and evaluation suites. Compatible with any host supporting AI Agent Skills, including Claude Code, Codex, Cursor, and Windsurf.
Not for: Not suited for: 1) Lightweight users only needing a quick copy rewrite (the full bundle introduces unnecessary overhead); 2) Teams unwilling to expose third-party API keys to session environments; 3) Unattended high-volume ad deployment or email sending without dry-run reviews (which directly triggers side effects on live production accounts); 4) Users seeking a single rigid methodology—the 50 skills span CRO/SEO/paid ads/retention with distinct tactical trade-offs requiring user judgment; 5) Users expecting fully autonomous marketing operations: it provides methodologies and tooling, not automated strategic decision-making.
Install Agent-Ready
① Verified Mirror 2026-09-15
Option A · Direct DownloadDownload marketing-skills.tar.gz
sha256:
Option B · Copy Guide for Agent27377979bea86ebe…Agents automatically download and verify mirrors from this JSON guide.
② Upstream GitHub
GitHub Repo ↗
Pinned to commit 5b2c000776.