Show HN：用于端到端视频广告营销活动的开源 AI 技能

让 AI Agent 创建营销视频和图片

一套开源技能，可让任何 AI Agent（Claude、Cursor、Codex、Hermes 等）生成端到端营销活动素材，包括 UGC 视频、广告视频、产品摄影等。

只需提供一张产品照片和一份简报。Agent 会提取产品信息，选择最优的 AI 图像与视频模型，挑选 AI 演员，并将生成的内容剪辑为可直接用于营销活动的成品素材。它可以生成任意时长的视频，同时保持演员和产品的高度一致性。

闭源 AI 营销 Agent 的开源替代方案。

UGC 视频

为这件 T 恤制作一支 UGC 视频。应当是一条“get ready with me”短视频。节奏明快，非常 Gen Z。 [产品照片]

卡通视频广告

制作一支风格化 3D 卡通视频广告，推广我的随行杯。要有趣、富有娱乐性。 [产品照片]

视频/嵌入内容：stanley_cartoon_ad_30s_newVO.mp4

产品广告视频

为这款能量饮料制作一支广告视频 [产品照片]

---

---

Claude 内置的一间完整制作工作室。你提供一张产品照片，以及一份概略（或详细）的简报。

随后，AI Agent 会规划视频 → 撰写脚本 → 生成 AI 演员 → 创建分镜 → 生成视频片段 → 将其拼接，并交付最终视频。

最重要的是：你始终掌握控制权。

---

将以下内容粘贴给你的编程 Agent（Cursor、Codex 等）：

Run `npx --yes github:SupercmoHQ/superCMO-skills --all`. Once it's installed, run the supercmo-setup skill to complete the setup.

如希望自行运行：

npx --yes github:SupercmoHQ/superCMO-skills --all   # every detected host

运行 supercmo-setup 技能，或让你的 Agent “帮我完成设置”。

希望在花钱前先预览？每次生成都支持 dry_run ——可查看准确请求和成本，且不会发起 API 调用。

---

从一张产品照片开始

从产品本身开始，比从空白提示词描述一支视频更快。

向 SuperCMO 提供一张产品照片，它就会围绕该产品制作 UGC 广告——由镜头前的创作者对你的产品进行评测、开箱、试穿或演示：

1. 上传产品照片并提供简短简报。

2. 它会分析产品、选择形式、挑选创作者并撰写脚本。随后向你展示创意概念，并在花费任何费用前等待你的批准。

3. 获批后，它会生成分镜、渲染每个片段、将片段拼接起来，并交付成品视频。

   "为这款产品制作一支 30 秒的开箱 UGC 广告。" [附上产品照片]

你得到的并不是一段随机视频片段，而是：

• 你始终掌控全局 ——在任何内容渲染前批准创作者、创意概念和脚本，并可在过程中修改任何部分。
• 内置一致性 ——通过分镜与锚点参考，在每个片段中维持同一演员和产品。
• 一支完整、抓人眼球的视频 ——按你要求的时长交付，可直接发布。

可在 Claude Code、Cursor、Codex、Hermes 中使用——即任何支持 Agent Skills 规范的 Agent。

---

你来执导，它来制作	SuperCMO 不会替你决定营销活动该传达什么——决定权在你。它处理底层制作工作：按你的简报写脚本、选角、拍摄、剪辑。它会在每一步与你确认，因此你可以在执行中途调整方向，而不必只能接受最终产出。
无需在工具间跳转	如今制作一支视频广告，通常要在多个工具间切换：一个写脚本、一个做图片、一个生成声音、一个生成视频，最后再在编辑器中拼接。SuperCMO 能根据一份简报，在同一处完成全部工作。
自动为你选择合适模型	每个 AI 图像/视频模型的工作方式都不同，需要不同的提示词。了解 Veo 需要何种提示词、Kling 又需要何种提示词，以及一开始该选择哪个模型，已是大部分工作。SuperCMO 的技能已掌握这些知识。你撰写简报；其余交给它处理。
每个镜头的一致性	做出一个好片段并不难。难的是在十个片段中让产品和人物面孔始终完全一致。SuperCMO 会在渲染前完成分镜，并将参考锚点锁定到每次生成中，使整支视频保持连贯。
包含剪辑	它不会只交付一段原始的 10 秒视频片段。它会生成媒体内容、拆分长片段、修剪素材、添加配音，并将所有内容拼接为完整成品。
按使用量付费，无订阅	大多数营销工具无论你是否使用都按月收费。SuperCMO 不这样做——你可以使用自己的供应商密钥（免费），或使用 SuperCMO 的密钥生成内容并按次付费。没有锁定。
在你已有的工作环境中运行	可在 Claude Code、Cursor、OpenAI Codex、Hermes、Openclaw，或任何支持 Agent Skills 规范的 Agent 内运行。你的密钥和文件始终保留在自己的机器上。

---

这些技能如何协同工作

你担任创意总监，而你的 Agent 使用 SuperCMO 技能来编排整个制作流程。

以下示例展示了这些技能如何在幕后串联，以执行一项复杂简报，例如：“为这个产品 [URL] 制作一支带配音的 1 分钟视频广告。”

[简报] "为这个产品 [URL] 制作一支带配音的 1 分钟视频广告。"
   │
   ├── 1. 分析 ──> 抓取 URL，以了解产品并获取产品图片。
   │
   ├── 2. 规划 ──> 将 60 秒简报拆分为四个 15 秒镜头。
   │                撰写脚本和镜头列表。
   │
   ├── 3. 图像 ──> 使用最佳图像模型生成产品的主参考图像。
   │
   ├── 4. 视频 ──> 生成 4 个视频片段。将锚点参考锁定至每个镜头中，
   │                以防产品在不同镜头中变形走样。
   │
   ├── 5. 音频 ──> 添加配音，并根据视频的准确时长规划脚本字数。
   │
   └── 6. 拼接 ──> 将完成的片段拼接为一个连续的 1 分钟视频文件。

营销制作流水线

SuperCMO 将一整家创意代理机构安装到你的文本 Agent 中。它不只是生成媒体内容；还会编排制作高转化广告所需的具体步骤：

• 产品分析 ——抓取你的 URL，提取品牌规范、材料和功能特点，避免 AI 模型虚构你的产品。
• 广告文案与脚本 ——将简报拆解为逐镜头广告脚本，并针对 Meta、TikTok 或 YouTube 等平台进行适配。
• 视觉生成 ——生成主参考图像，并将其锁定为锚点，以确保产品在每个视频帧中保持一致。
• 音频与配音 ——生成自然、有活力的配音，并使节奏与视频时长精准匹配。
• 后期制作 ——修剪素材、同步音频，并拼接最终 MP4 文件。

以下是安装到 Agent 中、用于运行该流水线的独立技能：

技能	描述
generating-ugc-videos	生成任意时长的 UGC 视频，如评测、开箱、试穿、教程等。它负责选择演员、撰写脚本，并为多个片段制作分镜，以锁定演员和产品的一致性。你可以指导每一步，在数分钟内得到连贯、可直接投放的广告。
generating-ad-videos	以你的品牌调性制作精致、电影感的产品广告——可以是无演员的产品展示，也可以是由演员推动故事的广告，且支持任意时长。它会为每个片段制作分镜，确保真实产品在不同剪辑间保持完全一致；当广告需要演员时，会撰写配音并进行选角；随后将内容拼接成一支成片。每一步都会先经你批准，再进行昂贵的渲染。
generating-cartoon-videos	为你的产品制作绘制风格的动画视频——可以是卡通、动漫、插画或绘画风格，且支持任意时长。你的产品可以保持与照片完全一致，也可以被绘制进特定风格中，由一组角色围绕它展开故事。它会确定一种美术风格、撰写故事、一次性绘制角色阵容以确保每个镜头使用相同角色，然后逐片段制作视频并叠加配音——在进行昂贵渲染前会先获得你的批准。
cloning-video-ads	用你的产品复刻你欣赏的视频广告。它会观看参考广告，分析使其奏效的镜头、节奏、摄影和配音，然后围绕你的产品重建相同结构——替换成你的产品、品牌和台词，同时保留原作的制作技艺。在进行任何昂贵渲染前，你可以批准方案；最终会交付一支完整视频。
generating-image-ads	为你的产品制作抓人眼球的图片广告——可以是一张图片，也可以是一组覆盖信息流所需各种版位比例的图片。它会在每个画面中让真实产品和品牌身份保持完全一致，并依据行业最佳实践，将你提供的优惠信息和主张整理为设计完善的广告。
generating-product-photos	围绕你的产品指导商业摄影拍摄，自动应用正确的灯光、构图、机位和道具，在 10 种以上的形式中交付影棚级产品照片。
researching-competitor-ads	分析竞争对手正在投放的广告，展示真正有效的内容——投放时间最长的广告、被悄然下线的广告，以及优胜者的共同点。视觉模型会完整观看承载信号的图片和视频广告——长期投放的、新近上线的、已被下线的——从头到尾。因此你可以了解每支广告的构成：开场钩子、节奏、产品何时出现、佐证和优惠。你还会获得所在品类尚未被采用的创意角度。
analyzing-own-ads	像分析竞争对手一样审计你自己的广告——从公开广告库中提取你投放的每支广告，并完整观看其创意内容。你将获得现有广告覆盖情况的全景图：正在使用哪些创意角度、钩子和形式，哪些广告维持时间最长，哪些被悄然停止以及停止速度如何，以及哪些优胜广告仍在支撑账户、却没有更新内容与之配合。
planning-campaigns	决定你的下一批广告应当是什么。它会了解你的品牌、产品、现有广告及竞争对手动向，然后按测试价值排序撰写创意概念——每项概念均完整说明一支广告、目标买家、所测试的假设、所依据的佐证，以及可用于制作它的技能。批准计划后，它会通过专业技能逐一制作你选择的概念。
onboarding-user	通过一个链接完成品牌设置。提供你的网站后，它会判断你是谁——你的语调、受众和销售内容——找到你的竞争对手，并将这些信息写入所有未来会话都能找到的位置。
supercmo-setup	通过几个简单步骤，让你从完成安装到开始生成内容。
adapting-formats	接收你已有的广告，并将其重制为所需的所有其他尺寸。它会先读取广告，确定哪些元素需要移动、需要用什么填补新尺寸带来的空间，并在花费任何费用前向你展示该方案；随后会围绕原稿重建每种尺寸，使产品、色板和文案保持可辨识的一致性，而不是被粗暴拉伸、裁切或加黑边。
analyzing-brand	从你的网站中分析品牌——颜色、字体和 Logo、表达方式、销售对象、所售产品，以及你支持的主张。网站没有实际说明的内容会被标记为未确认，而非凭空编造，因此用于规划和制作广告的技能会基于你的真实品牌开展工作。
analyzing-products	分析电商链接或照片，梳理产品的精确物理机制并提取干净的参考图像，无需人工研究或提取，即可构建高度准确的创意简报。
generating-ai-actors	选择具有鲜明特征、符合特定人口统计属性的 AI 演员，并将其作为可复用的身份参考，使你能够在跨多个图片和视频素材的整合营销活动中，稳定使用完全相同的品牌面孔。
generating-audio	将脚本转换为 aud

Enable AI agents to create marketing videos and images

Open-source skills that empower any AI agent (Claude, Cursor, Codex, Hermes, etc.) to generate end-to-end marketing campaigns - UGC videos, ad videos, product photography and more.

Just provide a product photo and a brief. The agent extracts the product details, selects the optimal AI image and video models, casts AI actors, and edits the generated content into finished, campaign-ready assets. It can produce videos of any length while maintaining perfect actor and product consistency.

The open-source alternative to closed AI marketing agents.

UGC video

create an ugc video for this tshirt. it should be a get ready with me reel. snappy. very genz. [product-photo]

Cartoon video ad

create a cartoon video ad in stylized 3D style, promoting my tumbler. make it fun and entertaining. [product-photo]

stanley_cartoon_ad_30s_newVO.mp4

Product ad video

make me an ad video for this energy drink [product-photo]

---

---

A full production studio inside Claude. You provide a product photo and a high level (or detailed) brief.

The AI agent then plans the video → writes the script → generates the AI actor → creates storyboards → generates the video clips → stitches them together, and delivers the final video.

The best part? You stay in charge all along.

---

Paste this to your coding agent (Cursor, Codex, …):

Run `npx --yes github:SupercmoHQ/superCMO-skills --all`. Once it's installed, run the supercmo-setup skill to complete the setup.

Prefer to run it yourself:

npx --yes github:SupercmoHQ/superCMO-skills --all   # every detected host

Run the supercmo-setup skill or ask your agent to "set me up".

Prefer a preview before spending? Every generation supports dry_run — the exact request and cost, no API call.

---

Start from a product photo

Starting from your product is faster than describing a video from a blank prompt.

Hand SuperCMO a single product photo and it builds a UGC ad around it - a creator on camera reviewing, unboxing, trying on, or demoing your product:

1. Upload a product photo and give a short brief
2. It analyzes the product, picks the format, casts the creator, and writes the script. Then shows you the concept and waits for your approval before spending a cent
3. Once approved, it generates the storyboards, renders each clip, stitches them together, and hands back the finished video

"Make a 30-second unboxing UGC ad for this product." [attach product photo]

What you get back is not a random clip. You get:

• You stay in charge - approve the creator, the concept, and the script before anything renders, and change any part along the way.
• Consistency built in - maintains the same actor and product across every clip, through the storyboard and anchor references
• A finished, scroll-stopping video - delivered at the length you asked for, ready to post

Works inside Claude Code, Cursor, Codex, Hermes - any agent that supports the Agent Skills spec.

---

|
---|---
You direct, it produces | SuperCMO doesn't decide what your campaign should say — you do. It handles the production work underneath: scripting to your brief, casting, shooting, editing. It checks in at every step, so you can redirect it mid-run instead of accepting whatever comes out the end.
No tool hopping | Making one video ad today means jumping between tools: script in one, images in another, voice in a third, video in a fourth, then stitching it in an editor. SuperCMO does it all from a single brief, in one place.
The right model, picked for you | Every AI image/video model works differently and needs different prompting. Knowing that Veo wants one kind of prompt and Kling another — and which to reach for in the first place — is most of the work. SuperCMO's skills already know. You write the brief; it handles the rest.
Consistency across every shot | One good clip is easy. Ten clips where the product and the face stay identical is the part that breaks. SuperCMO storyboards before it renders and locks a reference into every generation, so continuity holds across the whole cut.
Editing included | It doesn't just hand back a raw 10-second clip. It generates the media, splits long clips, trims footage, adds voiceover, and stitches everything into a finished asset.
Pay per use, no subscription | Most marketing tools charge monthly whether you use them or not. SuperCMO doesn't — bring your own vendor keys (free), or generate on SuperCMO's keys and pay per use. No lock-in.
Runs where you already work | Inside Claude Code, Cursor, OpenAI Codex, Hermes, Openclaw, or any agent that supports the Agent Skills spec. Your keys and files stay on your machine.

---

How the skills work together

You act as the creative director, and your agent uses SuperCMO skills to orchestrate the entire production pipeline.

Here is an example of how the skills chain together behind the scenes to execute a complex brief like: "Make a 1-minute video ad for this product [URL] with a voiceover."

[Brief] "Make a 1-minute video ad for this product [URL] with a voiceover."
   │
   ├── 1. Analyze ──> Scrapes the URL to understand the product and fetch product images.
   │
   ├── 2. Plan    ──> Breaks the 60s brief into four 15-second shots.
   │                  Writes the script and shot list.
   │
   ├── 3. Image  ──> Generates master reference images of the product 
   │                 using the best image models.
   │
   ├── 4. Video  ──> Generates the 4 video clips. Locks the anchor reference 
   │                  into every shot so the product doesn't shape-shift.
   │
   ├── 5. Audio   ──> Adds a voiceover, budgeting the script to match the 
   │                  exact length of the video.
   │
   └── 6. Stitch  ──> Stitches the finished clips together to create one 
                      continuous 1-minute video file.

The Marketing Production Pipeline

SuperCMO installs a full creative agency into your text-based agent. It doesn't just generate media; it orchestrates the specific steps required to make high-converting ads:

• Product Analysis - Scrapes your URLs to extract brand guidelines, materials, and features so the AI models don't hallucinate your product.
• Ad Copy & Scripting - Breaks the brief into a shot-by-shot ad script tailored for platforms like Meta, TikTok, or YouTube.
• Visual Generation - Produces master reference images and locks them as anchors so your product looks consistent in every video frame.
• Audio & Voiceover - Generates natural, high-energy voiceovers paced perfectly to your video duration.
• Post-Production - Trims footage, syncs audio, and stitches the final MP4.

These are the discrete skills installed into your agent to run that pipeline:

Skill	Description
generating-ugc-videos	Produces UGC videos of any length - like review, unboxing, try-on, tutorial and more. It casts the actor, writes the script, and storyboards multiple clips to lock in actor and product consistency. It lets you direct every step to get a cohesive, ready-to-run ad in minutes.
generating-ad-videos	Produces a polished, cinematic product commercial in your brand's voice — a no-actor product showcase or an actor-led story ad, at any length. It storyboards every clip to hold your real product identical across cuts, writes the voiceover and casts the actor when the ad calls for one, then stitches it into one ready-to-run spot you approve step by step before anything expensive renders.
generating-cartoon-videos	Produces a drawn, animated video for your product — cartoon, anime, illustrated or painted, at any length. Your product either stays exactly as photographed or is drawn into the style, and a cast of characters carries the story around it. It settles one art style, writes the story, draws your cast once so every cut is the same cast, then films it clip by clip and lays a voiceover over the top — with your approval before anything expensive renders.
cloning-video-ads	Recreates a video ad you admire with your own product. It watches the reference ad, works out the shots, pacing, camera and voiceover that make it work, then rebuilds the same structure around your product — swapping the product, brand and spoken lines for yours while holding the original's craft. You approve the plan before anything expensive renders, and it comes back as one finished clip.
generating-image-ads	Produces a scroll-stopping image ads for your product — one image, or a set in every placement ratio your feed needs. It keeps your real product and your brand identity perfectly consistent across every frame, and shapes the offer and claims you give it into a well designed ad, following industry best practices.
generating-product-photos	Directs a commercial photoshoot around your product, automatically applying the correct lighting, framing, camera angles and props to deliver studio-quality product shots across 10+ formats.
researching-competitor-ads	Analyzes what your competitors are advertising and shows you what is actually working — the ads that have run longest, the ones they quietly dropped, and what the winners have in common. A vision model watches the image and video ads that carry the signal — the long-runners, the ones just launched, the ones they dropped — the whole way through. So you learn how each one is built: the hook, the pacing, when the product lands, the proof and the offer. You also get the angles nobody in your category has taken.
analyzing-own-ads	Audits your own advertising the way you would a competitor's — pulling every ad you run from the public library and watching the creative the whole way through. You get the map of what you already cover: which angles, hooks and formats you run, which ads have held longest, what you quietly stopped and how fast, and which winners are still carrying the account with nothing newer built beside them.
planning-campaigns	Decides what your next ads should be. It reads your brand, your product, the ads you already run and what competitors are doing, then writes concepts ranked by what is worth testing — each a full description of one ad, the buyer it targets and the bet it tests, with the proof it rests on and the skill that would build it. Approve the plan and it builds the concepts you pick, one by one, through the specialist skills.
onboarding-user	Sets your brand up from one link. Give it your website and it works out who you are — your voice, your audience, what you sell — finds who you compete with, and writes both where every future session will find them.
supercmo-setup	Gets you from installed to generating in a few simple steps
adapting-formats	Takes an ad you already have and remakes it in every other size you need it in. It reads your ad first, works out what has to move and what fills the space the new shape opens up, and shows you that plan before it spends anything — then rebuilds each shape around the original so your product, palette and wording stay recognisably the same rather than being stretched, cropped or letterboxed into a mess.
analyzing-brand	Works out your brand from your website — your colours, fonts and logo, how you sound, who you sell to, what you sell, and the claims you stand behind. Anything your site doesn't actually say is written down as unconfirmed rather than invented, so the skills that plan and make your ads work from your real brand.
analyzing-products	Analyzes an e-commerce link or photo to map out your product's exact physical mechanics and extract clean reference images, building a highly accurate creative brief without any manual research or extraction.
generating-ai-actors	Casts a distinct, demographic-specific AI actor as a reusable identity reference, allowing you to reliably feature the exact same brand face across an entire multi-asset image and video campaign.
generating-audio	Converts scripts into audio. Lets you preview voice candidates from top models before generating. It then formats your script's numbers, dates, and URLs so the chosen AI model pronounces them accurately.
generating-images	Routes your brief to the best of 10+ SOTA image models (like Nano Banana, GPT Image), writes detailed prompts as per model specifications and generates multiple variations instantly.
generating-storyboards	Draws out your entire video concept clip-by-clip as static images, enforcing strict character and product continuity so you can approve camera angles and visual flow before committing to expensive video renders.
generating-videos	Routes your brief to the best of 10+ SOTA video models (like Veo, Seedance), writes detailed prompts as per model specifications and generates multi-clip stories that hold continuity across cuts, stitched into one file.
identifying-competitors	Finds who a brand competes with: searches the web for the alternatives buyers compare it against, proposes a shortlist with a source per name, and confirms the real ones for other skills to build on.
writing-ad-copy	Writes the text that runs alongside your ad — headline, primary text, description and call to action — tuned to the fields and best practices of each platform it runs on: Facebook and Instagram, Google Search, LinkedIn, TikTok, X and Reddit. It works one angle several ways and sizes each field so nothing gets cut off in the feed.
writing-video-prompts	Translates your scenes into the highly specific, detailed instructions that different video models require. It times the camera motion and steers around known AI rendering limits to match your exact visual brief.
writing-video-scripts	Drafts spoken scripts built around proven opening hooks designed specifically to stop the scroll. It sizes the text to fit your exact runtime and paces words across natural breath points for human delivery.

Every host loads the same skills + the same local MCP server - only the wiring differs. Prerequisite: uv on PATH — the MCP server runs via uvx, which provisions Python and the server for you (nothing to pip install). The root pyproject.toml / package.json are packaging scaffolding — you don't build anything by hand to run or contribute to the skills.

npx installer - registers the MCP server via each host's own mechanism (codex/claude mcp add where a CLI exists; a .cursor/mcp.json for Cursor) and places the skills. One command does every detected host:

npx --yes github:SupercmoHQ/superCMO-skills --claude               # Claude Code
npx --yes github:SupercmoHQ/superCMO-skills --cursor --project-dir .  # Cursor (per project)
npx --yes github:SupercmoHQ/superCMO-skills --codex                # Codex
npx --yes github:SupercmoHQ/superCMO-skills --all                  # every detected host

If the SuperCMO tools don't show up in your agent afterward, restart it once.

Claude Code plugin - an alternative to npx … --claude (use one, not both): the whole repo installs as one plugin, managed by Claude Code:

/plugin marketplace add SupercmoHQ/superCMO-skills
/plugin install supercmo@superCMO-skills

Codex plugin - an alternative to npx … --codex (use one, not both): installs the skills + MCP server as one plugin, managed by Codex:

codex plugin marketplace add SupercmoHQ/superCMO-skills
codex plugin add supercmo@superCMO-skills

Claude Cowork / Claude desktop - download supercmo-plugin.zip from the latest release, then Settings → Plugins → Upload local plugin.

Found us onskills.sh? npx skills add SupercmoHQ/superCMO-skills copies the skills but not the MCP server they call — so image_generate/video_generate won't exist yet. Install the server, then reload your agent:

npx --yes github:SupercmoHQ/superCMO-skills --all   # or --cursor --project-dir . · --codex · --all

Generation needs a key. Two ways — pick one:

Option A · Managed — one command

Generate on SuperCMO's keys, pay per use — no vendor signups, nothing to paste.

npx --yes github:SupercmoHQ/superCMO-skills login

Opens SuperCMO in your browser to sign in and authorize this device ; the key is written to ~/.supercmo/.env automatically. Buy credits in the web app when you're ready.

Option B · Bring your own keys — free

Bring your own vendor keys — requests go directly to the model vendor; nothing routes through SuperCMO.

One file, every host. The installer creates ~/.supercmo/.env; the MCP server loads your keys from there on any host (Claude Code, Cursor, Codex, …). Open it and add a key:

# ~/.supercmo/.env
WAVESPEED_API_KEY=your-key # image + video (start here) — https://wavespeed.ai
FAL_KEY=                   # image + video (alternative) — https://fal.ai
ELEVENLABS_API_KEY=        # voiceover (optional)        — https://elevenlabs.io
GEMINI_API_KEY=            # image/video analysis (opt.) — https://aistudio.google.com
FIRECRAWL_API_KEY=         # url extraction (optional)   — https://firecrawl.dev

• One key to start: WAVESPEED_API_KEY covers image + video and is what we recommend — users report better reliability there. FAL_KEY is fully supported too and serves the same models, so use it if that's the account you already have. With both set, WaveSpeed is used; SUPERCMO_MEDIA_PROVIDER=fal picks the other. The rest are optional — add one when you want that capability (voiceover, analysis, extraction).
• Prefer environment variables? Exporting the key (or putting it in your host's MCP config env block) also works and takes precedence — ~/.supercmo/.env is just the reliable default that also works for GUI-launched hosts that don't inherit your shell.
• Check what's set: ask your agent to run the setup_status tool (host-agnostic, no path needed).

These skills run inside your agent. Exactly what happens:

• Open source & inspectable. Every skill, script, and the MCP server lives in this repo under Apache-2.0 — read, diff, or pin before you run it.
• Your keys stay local (BYOK). With your own vendor keys, they live in ~/.supercmo/.env (or your host's MCP config env / your shell); the server reads them from its process environment and requests go directly to the vendor — nothing routes through SuperCMO. With the managed key, requests go to SuperCMO's proxy (billed to your credits) — you never hand us a vendor key.
• The MCP server is local + minimal. A stdlib-only Python package (supercmo-skills on PyPI, source in scripts/supercmo_skills/mcp/), fetched and run on demand via uvx supercmo-skills@<version> — it runs on your machine, launched by your host, and starts only when your host enables the plugin.
• Dry-run everything. Generation tools support dry_run - a free preview of the exact request (keys masked), no API call.

Found something off? Open an issue.

SuperCMO sends anonymous, opt-out usage counts from the MCP server so we can see which tools get used and prioritize. Full details in TELEMETRY.md.

• What we send: the tool name, whether it succeeded, how long it took, versions (OS / Python / SuperCMO), and a random install id. Nothing else.
• What we NEVER send: your prompts, tool arguments, generated media, file paths, keys, hostname, username, or IP address.
• Turn it off (any one): SUPERCMO_TELEMETRY=false, DO_NOT_TRACK=1, or DISABLE_TELEMETRY=1. It also honors Claude Code's CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC.
• See what would be sent: run with SUPERCMO_TELEMETRY=log - prints each payload instead of sending.
• Events go to our own endpoint (api.getsupercmo.ai), never a third-party analytics host. The random install id is never linked to any account.

Community

Questions, ideas, or something you built? Open an issue or discussion.

New skills and improvements welcome - see CONTRIBUTING.md. In short:

1. Create skills/<your-skill>/ (folder name = the name in frontmatter), using skills/generating-images/ as the reference layout.
2. Keep SKILL.md short; push detail into references/, deterministic work into scripts/ (stdlib-only where possible, BYO-keys from env, --dry-run on anything that mutates).
3. Validate - CI runs the same on every PR:

python3 scripts/quick_validate.py       # structural + strict-YAML frontmatter (blocking)
python3 scripts/listing_gate.py         # scripts compile + --dry-run gates (blocking)
python3 scripts/check_shared_client.py  # no raw vendor HTTP - the brokering seam (blocking)
python3 scripts/check_catalog_sync.py   # provider-key catalog single-sourced (blocking)

Apache-2.0 - see LICENSE and NOTICE.

The skill files, scripts, and MCP server in this repository are Apache-2.0. The hosted SuperCMO product (getsupercmo.ai) is a separate service governed by its own terms.