aiskilloftheweek/idea-extractor
Analyzes long-form content (transcripts, articles, interviews, webinars, podcast notes) and extracts the 3 best content ideas for any publishing platform — LinkedIn, Substack, blog, newsletter, podcast, YouTube, or other. Use this skill whenever the user provides a transcript, long-form content, or any extended text and wants to identify the best post ideas, extract content angles, mine a transcript for social content, or turn a recording/article into post ideas. Trigger on phrases like "extract ideas from this", "find post ideas in this transcript", "what can I post from this", "mine this content for ideas", "turn this transcript into posts", "help me repurpose this content", or any time the user pastes a long text and asks what to do with it for content purposes — regardless of platform. Always use this skill when repurposing long-form content into post ideas, even if the user phrases it casually.
npx skills add https://github.com/aiskilloftheweek/claude-ai-skill-of-the-week --skill idea-extractor
Analyze long-form input content and extract the 3 best content ideas, tailored to the user's target platform, voice, positioning, and audience.
Before extracting ideas, you need to understand who is creating the content, for whom, and where it will be published.
Check available context first: Look for any skill files, memory, or documents that contain information about the user's:
Always ask the target platform, even if other context is available — this is essential to shape the format and angle of each idea:
> "Which platform(s) are these ideas for? (e.g., LinkedIn, Substack, blog, newsletter, YouTube, podcast, Instagram...)"
If other context is also missing, ask the following before proceeding:
Do not proceed to Step 2 until you have the target platform confirmed and enough context to personalize the output.
Read the entire long-form input carefully. While reading, identify:
From all the ideas you identified, select the top 3 based on these criteria:
Present the 3 selected ideas using this exact XML structure:
<post_ideas>
<topic1>
<title>[Concise, compelling title for the post]</title>
<description>[Explain what the post is about, why it would resonate with the target audience, and what angle or hook to use. Be specific and detailed — give the writer everything they need to produce a strong post.]</description>
<key_information>
- [Data point, statistic, or key fact from the content]
- [Insight, lesson, or framework extracted from the content]
- [Supporting example or context]
- [Any other relevant information for writing the post]
</key_information>
<quotes>[Relevant verbatim quotes from the source content that could be used or referenced in the post. Leave empty if none.]</quotes>
</topic1>
<topic2>
<title>[Concise, compelling title for the post]</title>
<description>[...]</description>
<key_information>
- [...]
</key_information>
<quotes>[...]</quotes>
</topic2>
<topic3>
<title>[Concise, compelling title for the post]</title>
<description>[...]</description>
<key_information>
- [...]
</key_information>
<quotes>[...]</quotes>
</topic3>
</post_ideas>
Take aiskilloftheweek/idea-extractor from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.