An entire marketing department as one Claude skill. 14 modules: audits scored 0-100, an 18-tactic hook engine, copy graded before you see it, ad diagnosis, GEO, email, launches, pricing. Works in Claude, Codex and Cursor.
Yuzzyuk/marketing-os 排名第 6,766，有 547 个星标。
同类别比较
Yuzzyuk/marketing-os 排名第 6,766，此处有 8 个同类条目；bublint/ue5-llama-lora 排名第 6,767。
同类别比较| 名称 | 排名 | 指标 |
|---|
| samyost1/3dicon | 6,740 | - 语言:
- Python
- 星标数:
- 560
- 分叉数:
- 48
|
|---|
| QuantaAlpha/RepoMaster | 6,750 | - 语言:
- Python
- 星标数:
- 554
- 分叉数:
- 72
- 简介:
- RepoMaster: The open-source AI agent that masters GitHub. It turns any code repository into a powerful tool, achieving a new level of autonomous task-solving. An open alternative to Claude-Code.
|
|---|
| ibrahimcetin/reins | 6,757 | - 语言:
- Dart
- 星标数:
- 551
- 分叉数:
- 91
- 简介:
- Local AI chat for iPhone, iPad and Mac. Connect Ollama, LM Studio or any OpenAI-compatible server, or run MLX models on-device. Includes tools and web search.
|
|---|
| ImKKingshuk/LockKnife | 6,760 | - 语言:
- Python
- 星标数:
- 549
- 分叉数:
- 121
- 简介:
- LockKnife: The Ultimate Android Security Research Tool. A unified TUI workspace and headless CLI for deep Android security research, built for researchers and hackers. Powered by Python orchestration and a Rust-accelerated core, enabling AI agent–driven hacking, credential recovery/cracking, APK analysis, intelligence gathering, runtime inspection.
|
|---|
| bublint/ue5-llama-lora | 6,767 | - 语言:
- Python
- 星标数:
- 547
- 分叉数:
- 44
- 简介:
- A proof-of-concept project that showcases the potential for using small, locally trainable LLMs to create next-generation documentation tools.
|
|---|
| makerever/rever | 6,772 | - 语言:
- TypeScript
- 星标数:
- 546
- 分叉数:
- 36
- 简介:
- Rever AI - powered finance automation that elevates control, unifies systems, and enables foresight for modern CFOs.
|
|---|
| omnichar/OmniChar | 6,773 | - 语言:
- Python
- 星标数:
- 546
- 分叉数:
- 73
- 简介:
- The open .char format for consistent characters. Face, body, wardrobe & voice in one file, consistent across every model.
|
|---|
| rakuri255/UltraSinger | 6,774 | - 语言:
- Python
- 星标数:
- 546
- 分叉数:
- 66
- 简介:
- AI based tool to convert vocals lyrics and pitch from music to autogenerate Ultrastar Deluxe, Midi and notes. It automatic tapping, adding text, pitch vocals and creates karaoke files.
|
|---|
关键参数与规格
关键参数与规格| 参数 | 值 |
|---|
| 类别 | Applications |
|---|
| 子类别 | Productivity,Specialists |
|---|
| 关键词 | Automation, Marketing, Structured outputs, Writing |
|---|
| 星标数 | 547 |
|---|
| 分叉数 | 108 |
|---|
| 贡献者 | 0 |
|---|
| 创建日期 | 2026-08-17 |
|---|
| 更新日期 | 2026-10-15 |
|---|
| 1 天星标数 | 1 |
|---|
| 7 天星标数 | 9 |
|---|
| 排名 | 6,766 |
|---|
| 上次排名 | 6,946 |
|---|
| 排名变化 | 180 |
|---|
| 7 天星标排名 | 4,318 |
|---|
| 上次 7 天星标排名 | 4,087 |
|---|
| 7 天星标排名变化 | -231 |
|---|
| 总星标排名 | 18,033 |
|---|
| 上次总星标排名 | 18,034 |
|---|
| 总星标排名变化 | 1 |
|---|
README 摘录
Product Hunt | | positioning | Positioning, offer design, pricing strategy | | competitive | Competitor teardowns from public sources | | app-store | ASO diagnosis + metadata + screenshot sequences | | analytics | Honest data reads, proper test design | | + slop-patterns | The AI-tell catalogue, run on all prose | | + rubrics/specs | Scoring bands, engine playbooks, store field rules | Ads briefs don't stop at paper: if an ad-generation MCP is connected (e. Install **Claude Code (one command):** **Claude (web/desktop):** Settings → Skills → + → upload from the latest release. **Manual / Codex / Cursor / other
agents:** copy into your agent's skills directory (e. First five minutes Copy → , fill it in, drop it in your project root or . Design principles **Progressive disclosure**: the router loads first (~2k tokens); each task loads only its module. Fourteen modules cost nothing until used. **Subagent-native**: multi-dimensional work (audits, teardowns, variant generation) fans out in parallel when the host supports it. **Honesty spine**: no invented proof, no fake precision, no winners declared on ten conversions, and every score labeled
as the heuristic it is.
常见问题
记录了多少星标和分叉？
Yuzzyuk/marketing-os 有 547 个星标和 108 个分叉。
仓库排名是多少？
Yuzzyuk/marketing-os 排名第 6,766。
哪个仓库排名最接近？
Yuzzyuk/marketing-os 与 bublint/ue5-llama-lora 比较，后者排名第 6,767，有 547 个星标。
Marketing OS
One skill. Fourteen modules. The full surface a marketer touches.
Built by tearing down the most-starred marketing skill repos on GitHub —
from 44K-star collections to 100-star craft pieces — keeping the structures
that worked, and adding the three things every one of them lacked: scores on
everything, artifacts instead of advice, and an explicit "what I couldn't
determine" in every report.
What's inside
| Module |
Covers |
| audit |
Website/funnel audit → weighted 0-100 + prioritized fixes |
| geo |
AI-search citability (ChatGPT, Perplexity, AI Overviews) |
| copy |
Generate 15-20 variants → expert-panel score → de-slop |
| hooks |
18-tactic hook engine → hook matrix → diagnostic funnel |
| paid-ads |
Concept-level fatigue diagnosis → ranked production brief |
| email |
Welcome/nurture/launch sequences, written in full |
| social |
LinkedIn/X posts that survive the feed |
| launch |
Launch playbook incl. Product Hunt |
| positioning |
Positioning, offer design, pricing strategy |
| competitive |
Competitor teardowns from public sources |
| app-store |
ASO diagnosis + metadata + screenshot sequences |
| analytics |
Honest data reads, proper test design |
| + slop-patterns |
The AI-tell catalogue, run on all prose |
| + rubrics/specs |
Scoring bands, engine playbooks, store field rules |
Ads briefs don't stop at paper: if an ad-generation MCP is connected
(e.g. Arcads), the skill generates the briefed assets directly —
diagnose → brief → produce, in one session.
Install
Claude Code (one command):
/plugin marketplace add Yuzzyuk/marketing-os
/plugin install marketing-os@marketing-os
Claude (web/desktop): Settings → Skills → + → upload marketing-os.zip
from the latest release. Requires code execution enabled.
Manual / Codex / Cursor / other agents: copy skills/marketing-os/ into
your agent's skills directory (e.g. ~/.claude/skills/marketing-os/).
First five minutes
Copy brand-context.template.md → brand-context.md, fill it in, drop it in
your project root or .claude/. Every module reads it. Skip this and the
output is competent and interchangeable — the one thing marketing can't be.
Design principles
- Progressive disclosure: the router loads first (~2k tokens); each task
loads only its module. Fourteen modules cost nothing until used.
- Subagent-native: multi-dimensional work (audits, teardowns, variant
generation) fans out in parallel when the host supports it.
- No executable code. Pure markdown. Nothing to review before trusting it.
- Honesty spine: no invented proof, no fake precision, no winners declared
on ten conversions, and every score labeled as the heuristic it is.
MIT.