content-repurposing

Overview

What the content-repurposing skill does

The content-repurposing skill helps you atomize a single long-form piece of content into many smaller, channel-ready assets. It is built around the inference.sh CLI and is designed for:

• Turning a blog post, podcast, video, or whitepaper into multiple derivative pieces
• Generating visual quote cards for social media from pull-quotes
• Supporting a repeatable content marketing workflow across platforms

This skill follows a content pyramid approach where one core source asset expands into medium-form and short-form outputs such as newsletters, social threads, posts, and visual snippets.

Who this skill is for

Use content-repurposing if you are:

• A content marketer or social media manager who wants structured AI help to break down long-form content
• A blogger, podcaster, or video creator looking to squeeze more distribution value out of each asset
• A marketing team experimenting with AI-assisted workflows using a CLI rather than ad‑hoc web tools

It is a strong fit when you already produce long-form content and want a systematic way to turn it into:

• Blog-to-thread or blog-to-carousel outlines
• Podcast-to-blog or summary structures
• Video-to-quotes and visual quote cards

When content-repurposing is or is not a good fit

Good fit when:

• You are comfortable running commands in a terminal
• You use or are willing to install the inference.sh (infsh) CLI
• You want reusable workflows and scripts, not a one-off copy-paste tool

Not ideal when:

• You need a fully graphical, non-technical interface with no CLI at all
• You only publish short updates and do not work with long-form source content
• You are looking for a generic AI writing assistant rather than a content atomization workflow

How to Use

1. Prerequisites and installation

The content-repurposing skill assumes you are using the inference.sh skills ecosystem and CLI.

1. Install the skill into your skills workspace (for example, an Agent Skills–compatible setup):

   npx skills add https://github.com/inferen-sh/skills --skill content-repurposing
   

2. Install the inference.sh CLI (infsh).

   The skill’s SKILL.md references the CLI install instructions:

   • Open: https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md
   • Follow the platform-specific steps there to install infsh and make sure it is on your PATH.

3. Log in to inference.sh once the CLI is installed:

   infsh login
   

After this, the content-repurposing workflows can call AI models via the infsh CLI.

2. Core workflow: from long-form content to assets

The core idea behind content-repurposing is the content pyramid:

• Start from a long-form source (blog post, podcast transcript, video transcript, whitepaper)
• Derive medium-form assets (newsletter sections, LinkedIn posts, email content)
• Break those into short-form snippets (tweets, hooks, captions, quotes, and images)

While the repository preview focuses on quote card generation, the skill description makes clear it is intended for:

• Blog-to-thread
• Blog-to-carousel
• Podcast-to-blog
• Video-to-quotes
• Other “one-to-many” content recycling recipes

These recipes are organized as guides in guides/content/content-repurposing within the repo, and you can adapt them to your own process.

3. Example: generating a quote card from a pull-quote

A concrete example in SKILL.md shows how to use infsh with an image model to create a social quote card.

1. Identify a strong pull-quote from your blog post or transcript.

2. Run a model via infsh to generate a background image suitable for a quote card:

   infsh app run falai/flux-dev-lora --input '{
     "prompt": "minimal quote card design, dark navy background, large white quotation marks, clean sans-serif typography space, modern professional design, social media post format",
     "width": 1024,
     "height": 1024
   }'
   

3. Use the generated image as the base for your quote graphic, then overlay your selected text in your design tool of choice.

This pattern can be extended, for example, by changing the prompt, dimensions, or target model to match different social platforms or brand styles.

4. Understanding the file layout

After installing content-repurposing, inspect these key files to understand and customize the workflow:

• SKILL.md – High-level description of the skill, requirements, and example usage snippets. This is the best starting point.
• guides/content/content-repurposing/ – Guide content that explains content atomization concepts such as the content pyramid and practical conversion recipes for different formats.

Depending on how you integrate skills, you may also see or use:

• metadata.json or similar config files in the broader inferen-sh/skills repo (if present in your local checkout) to understand how the skill is registered and described.

5. Adapting content-repurposing to your own workflow

Treat the repository as a reference implementation rather than a strict template. To adapt it:

• Swap in your own models: If you prefer different image or text models via inference.sh, adjust the infsh app run commands.
• Change prompts and parameters: Tailor the prompt text, image size, and aspect ratio to fit the social platforms and brand guidelines you use most.
• Integrate with your scripts: Wrap the example CLI calls in your own shell scripts or automation pipelines (for example, Bash scripts that pull content from your CMS, process it, and save generated assets).

The content-repurposing skill is particularly useful when you are building a repeatable content marketing workflow that you can run for every new long-form asset.

FAQ

Is content-repurposing a standalone app or a skill I install?

content-repurposing is a skill within the inferen-sh/skills repository. You add it to your skills environment using:

npx skills add https://github.com/inferen-sh/skills --skill content-repurposing

It is not a standalone desktop or web application; it is meant to be used programmatically or via terminal workflows.

Do I need the inference.sh CLI to use content-repurposing?

Yes. The skill’s SKILL.md explicitly lists Bash (infsh) as the allowed tool and provides commands that assume the inference.sh CLI (infsh) is installed.

You must:

1. Install infsh using the instructions at:

   https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md

2. Log in with infsh login before running the example commands.

Can content-repurposing handle more than quote cards?

The provided example focuses on quote card generation using an AI image model, but the skill is conceptually designed for broader content atomization:

• Blog-to-thread and blog-to-carousel flows
• Podcast-to-blog conversions
• Video-to-quotes workflows

The detailed guides and recipes for these transformations live under guides/content/content-repurposing/. You can follow those patterns and extend them with your own prompts and scripts.

Does content-repurposing write copy for me?

The repository excerpt specifically shows visual generation (quote cards) via infsh. While the high-level description references multiple content formats, you should treat text-generation patterns as guides to be implemented with your own model choices and prompts.

In practice, you would:

• Use inference.sh-compatible text models to outline threads, posts, or summaries
• Use the content pyramid and recipes as a workflow blueprint

Check the guides in the repo to see the latest recommended prompts and command patterns.

Is this skill suitable for non-technical marketers?

It is best suited to technical or hybrid marketers comfortable with a terminal. You do not need deep engineering skills, but you should be able to:

• Install a CLI
• Run example Bash commands
• Adjust prompts and parameters in infsh calls

If you need a purely graphical tool, you may want to pair with a teammate who can set up the commands and scripts for you.

Where can I see or edit the underlying configuration?

After adding the skill, open the files in your local checkout of the inferen-sh/skills repo:

• Start with SKILL.md for the overview and core command examples
• Explore guides/content/content-repurposing/ for the conceptual model and step-by-step recipes

You can copy these patterns into your own repository or internal tooling, then adjust paths, prompts, and model choices to match your content stack.