kokasexton/awesome-marketing-repos instruction manual

1. Purpose, scope, and Claude Code classification

Purpose

`kokasexton/awesome-marketing-repos` is a public, curated directory of open-source GitHub repositories for marketing teams. It is a reading and discovery resource, not an executable marketing product. Its documented function is to help readers compare tools for AI-assisted marketing, automation, analytics, SEO, email, social media, lead generation, design, advertising, experimentation, and related infrastructure.

The main interface is `README.md`. Each catalog entry normally provides a repository link, difficulty label, intended audience, short explanation, and a GitHub stars badge; some also show a last-commit badge. Introductory “Start here if” notes explain which kinds of users should browse each category. The Editor’s Picks section highlights eight repositories that the curator considers especially useful, while Recently Added points to new entries and their destination categories.

Relationship to Claude Code

**Classification: catalog; install mode: unclassified; confidence: high.** The repository relates to Claude Code by cataloging the official Claude Code repository and numerous Claude Code skills, actions, workspaces, MCP-backed tools, and orchestration systems. It does not document its own Claude Code plugin, skill, hook, MCP server, command, context injection, or installation mechanism. Therefore it should not be classified as a Claude Code extension or standalone companion application. Its reliable mechanism is discovery: readers follow links to separately maintained projects and evaluate those projects’ own first-party instructions.

The metadata identifies the default branch as `main`, enables issues and pull requests, and provides no repository license. Do not assume that the catalog’s contents, or every linked project, share a common license.

2. Reading and navigating the catalog

Navigation workflow

Begin with **Editor’s Picks** when you want a short list. The eight highlighted choices cover workflow automation (`n8n`), product analytics and experimentation (`PostHog`), marketing automation (`Mautic`), social scheduling (`Postiz`), CRM (`Twenty`), feature experiments (`GrowthBook`), data ingestion (`Airbyte`), and AI marketing skills (`marketingskills`). These are editorial recommendations, not measured comparisons supplied by this repository.

Use **Recently Added** to find entries newly incorporated in the stated August 2026 update. This area includes Claude Code and Agent Skill resources, SEO tools, GTM systems, growth tools, AI frameworks, MCP servers, and another Claude Code ecosystem list. Follow the category link shown after an item to read its normal entry and audience guidance.

The **Contents** list is the category map. In the supplied README evidence it names Editor’s Picks, Recently Added, 14 topical catalog sections, Marketing Stack Templates, Open-Source Alternatives, Browse by Tag, an AI Marketing Prompt Library, an Alphabetical Index, and Contributing. The supplied README body is truncated during Awesome Lists, so the later named sections cannot be described beyond their table-of-contents labels.

Within a topical section, read the “Start here if” paragraph first, then compare entries by:

1. **Difficulty** — Beginner, Intermediate, or Advanced.
2. **Best for** — the intended user or use case.
3. **Description** — the marketing problem and claimed role.
4. **Badges** — dynamic GitHub stars and, where present, last-commit information.

A listing is a pointer, not installation documentation. Open the linked first-party repository before adopting or running anything.

3. AI tools, agent infrastructure, LinkedIn, and email

AI Marketing Tools & Agents

This category is for AI-assisted copy, experiments, competitor research, content production, and repeatable workflows. It includes Claude Code skill collections such as `coreyhaines31/marketingskills`, `OpenClaudia/openclaudia-skills`, and `rampstackco/claude-skills`; agent suites with subagents or live execution; the official `anthropics/claude-code`, `anthropics/skills`, and `anthropics/claude-code-action` repositories; cross-CLI knowledge and skill packs; AI writing and lead-generation applications; and a Vercel Labs marketing-agent-team template. Use the stated difficulty and audience labels to distinguish simple skill packs from advanced MCP-backed or CI execution systems.

AI Frameworks & LLM Infrastructure

This section supports teams building custom pipelines rather than merely selecting prompts. It covers code frameworks (`LangChain`), visual builders (`Langflow`, `Flowise`), multi-agent systems (`CrewAI`, `AutoGen`), memory and RAG layers (`mem0`, `Quivr`, `Graphiti`), full agent platforms (`Dify`, `Agno`), automation (`n8n`), model routing (`LiteLLM`), MCP servers, current-document retrieval (`Context7`), Obsidian connectivity, and Claude Code orchestration. The README suggests LangChain for code-oriented work and n8n for visual workflows.

LinkedIn Tools & Outreach

Use this category for lead discovery, profile or job data, outreach, publishing, and multichannel GTM. Entries range from scrapers and an unofficial API wrapper to Apify clients, Claude Code/Codex content skills, and a skills-plus-MCP GTM system. The repository explicitly warns that LinkedIn’s Terms of Service restrict automated collection; review those terms before production use.

Email Marketing & Automation

This section covers self-hosted campaign automation (`Mautic`), omnichannel journeys, newsletters (`listmonk`), transactional and marketing email (`Plunk`), surveys and follow-up (`Formbricks`), and Markdown-based release-note or documentation publishing (`MkDocs`).

4. Analytics, SEO, social media, CRM, and lead generation

Analytics & Attribution

Use these listings to measure website, product, campaign, and behavioral performance. Options include all-in-one product analytics (`PostHog`), privacy-oriented web analytics (`Plausible`, `Umami`), full data ownership (`Matomo`), enterprise event infrastructure (`Snowplow`), a Segment-compatible CDP (`RudderStack`), warehouse transformation (`dbt`), dashboards (`Metabase`), source ingestion (`Airbyte`), and AI market-map research. The section’s guidance presents PostHog as the broad single-platform choice, Plausible as a privacy-friendly website option, and Airbyte as a warehouse foundation.

SEO & Content Optimization

This area contains an open Semrush/Ahrefs alternative, an SEO-tools directory, a Claude Code workspace for long-form content, a parallel Claude Code SEO skill suite, Search Console analysis through MCP, a website specification covering accessibility/security/SEO/agent readiness, a GEO workflow, and a local CLI plus MCP server offering SEO audits from user-provided data. Treat claims such as “only SEO skill” or “de-facto checklist” as catalog descriptions, then verify them at the linked source.

Social Media & Community

Browse this category for multi-platform scheduling (`Postiz`, `Mixpost`), OAuth social login, n8n posting and monitoring workflows, campaign planning (`Plane`), X/Twitter research (`Twint`), username discovery (`Sherlock`), and global news or trend monitoring. The section distinguishes publishing platforms from research and infrastructure tools.

Lead Generation & CRM

This category covers a modern self-hosted CRM (`Twenty`), integrated marketing/sales/support (`Erxes`), lead-capture backends (`Supabase`), usage billing (`Lago`), forms and surveys, AI-qualified lead lists, custom crawlers, triggered email, autonomous LinkedIn outreach, and a Python lead-generation toolkit. Choose based on whether the need is system-of-record management, data capture, enrichment, outreach, or supporting infrastructure.

5. Creative work, advertising, growth, automation, and learning

Design & Creative Assets

This section spans design tools and the technology used to ship marketing assets. It lists Penpot as an open Figma alternative; Next.js, Astro, and Tailwind CSS for sites and landing pages; coding-agent video editing; a Claude Code promotional-video skill; text-to-video production; and AI-generated presentations. The catalog does not supply asset-generation commands, so use each linked project’s own documentation.

Advertising & Paid Media

The paid-media section combines campaign data management (`NocoDB`), dashboards (`Metabase`, `Superset`), ad-source ingestion (`Airbyte`), and a Claude Code skill with MCP setup for Google Ads analysis. This arrangement supports a pipeline from collecting ad data to reporting and AI-assisted auditing; it does not claim that every listed tool directly modifies live ad accounts.

Growth Engineering & Experimentation

Use this category for A/B testing, feature flags, controlled rollouts, commerce infrastructure, and autonomous growth loops. `GrowthBook` and `PostHog` are the highlighted experiment platforms; `Flagsmith` and `Unleash` focus on feature management; `Medusa` supports commerce; and `growth-lab` is described as researching, executing, reviewing, and improving campaigns.

Marketing Automation Platforms

This section lists broad campaign automation (`Mautic`), visual integration (`n8n`), Zapier-style alternatives (`Activepieces`, `Automatisch`), developer-oriented background workflows (`Trigger.dev`), a multi-platform SMM toolkit, and the Google Workspace CLI. Difficulty labels distinguish low-code workflow assembly from engineering-heavy infrastructure.

Learning Resources & Prompts

Use this area for prompt collections, prompt-engineering education, weekly AI research, prompt evaluation and red teaming, and first-party API recipes from OpenAI and Anthropic. These are learning or testing resources rather than complete marketing automation platforms.

The supplied **Awesome Lists & Curations** portion points to broader marketing, AI-marketing, SEO, and Claude Code ecosystem directories for further discovery.

6. Selecting entries safely and interpreting labels

Selection method

First define the marketing outcome: for example, private analytics, campaign automation, SEO auditing, a self-hosted CRM, or a Claude Code skill. Choose the matching category, read its “Start here if” note, then shortlist entries whose “Best for” label matches your team.

Interpret difficulty using the contribution guide’s exact definitions:

• **`[Beginner]`** — usable without writing code; UI-based or no technical setup required.
• **`[Intermediate]`** — needs configuration, basic command-line work, or light scripting.
• **`[Advanced]`** — needs engineering expertise, infrastructure, or substantial setup.

These labels describe expected setup, not security, maintenance quality, legal suitability, or total operating cost. Stars and last-commit badges are discovery signals only. The contribution policy expects public GitHub access, activity within 18 months, a real marketing use case, meaningful documentation, and normally at least 50 stars, although exceptionally unique or new projects may be accepted.

Before using a linked repository, independently inspect its first-party README, license, maintenance state, deployment model, required credentials, data handling, and platform terms. This is especially important for social scraping, autonomous outreach, account-connected MCP servers, ad-account execution, and tools that send customer data to model providers. The catalog itself documents no common security review or compatibility test.

Do not infer that “open-source alternative” means feature parity with a named commercial service. Likewise, descriptions such as “replace HubSpot,” “replace Semrush,” or “gold standard” express catalog positioning. Validate features against your own requirements.

Finally, note evidence limits: the repository metadata reports no license and the supplied README ends partway through Awesome Lists. No installation, test, build, release, or runtime command for the catalog is documented.

7. Contributing a repository by pull request

Eligibility checklist

Submit a public GitHub repository that is actively maintained, solves a real marketing problem, has meaningful documentation, and generally has at least 50 stars. A unique new project may be considered below that threshold. Do not submit paid or closed-source tools, repositories inactive for more than 18 months, generic developer tools without a clear marketing use case, or undisclosed self-promotion.

The preferred workflow is to fork the repository, then clone your fork using the documented commands. Replace `YOUR-USERNAME` before copying:

git clone https://github.com/YOUR-USERNAME/awesome-marketing-repos.git
cd awesome-marketing-repos

Add the entry to the most relevant section of `README.md` using the documented structure:

- **[owner/repo](https://github.com/owner/repo)** `[Beginner/Intermediate/Advanced]` `Best for: audience description` — One clear sentence describing what it does and why marketers care.
  ![Stars](https://img.shields.io/github/stars/owner/repo?style=flat-square) ![Last Commit](https://img.shields.io/github/last-commit/owner/repo?style=flat-square)

Select exactly one real difficulty label rather than retaining the slash-separated placeholder. Keep “Best for” short and specific. The guide gives examples such as `Best for: replacing HubSpot`, `Best for: non-technical marketers`, `Best for: agencies managing multiple clients`, and `Best for: B2B demand gen teams`.

Then add the repository to the Alphabetical Index and Recently Added sections, and update the tag section when relevant. Submit a pull request titled:

Add: owner/repo — One-line description

If you created the submitted project or are affiliated with it, disclose that relationship in the pull-request description. The contribution guide welcomes creator submissions when the conflict is transparent.

8. Issues, formatting rules, corrections, and conduct

Contributing without a pull request

Open an issue at:

https://github.com/kokasexton/awesome-marketing-repos/issues/new

For a new repository request, provide its URL; select a category or propose a new one; choose Beginner, Intermediate, or Advanced; identify who benefits; explain the marketing use case in one paragraph; and report current stars and last-commit information. The documented categories include AI marketing, AI infrastructure, LinkedIn, email, analytics, SEO, social, lead generation/CRM, design, advertising, growth, automation, learning, and awesome lists.

To report a stale or incorrect entry, provide the repository URL and identify whether it is archived, inactive for more than 18 months, a broken link, misleadingly described, or in the wrong category. Add details that let maintainers verify the report. For a new section or feature, describe the proposal, why it is valuable, and example content.

Badge formatting

Every proposed listing should include both badges in this exact documented form, replacing the uppercase placeholder:

![Stars](https://img.shields.io/github/stars/OWNER/REPO?style=flat-square) ![Last Commit](https://img.shields.io/github/last-commit/OWNER/REPO?style=flat-square)

Descriptions should be one clear sentence explaining both function and marketing relevance. Put the entry in the most relevant section and keep indexes, recent additions, and applicable tags synchronized.

Review and conduct

Contributions are reviewed on merit. Be respectful. If a submission is rejected and you disagree, open a discussion issue rather than repeatedly resubmitting or escalating. Questions may be filed as an issue or directed to the repository owner, `@kokasexton`, on GitHub.

Because no automated validation workflow is documented in the supplied files, manually check links, labels, badge paths, category placement, indexes, and disclosure before submission. Do not add undocumented execution instructions for linked tools; link readers to those projects’ first-party documentation instead.