Can I use this recipe in production?

Yes. Every recipe is production-tested with error handling, logging, and deployment guidance.

Which LLM providers are supported?

Recipes support OpenAI, Anthropic Claude, Google Gemini, and open-source models via a unified interface.

How do I customize these recipes?

Each recipe includes a configuration section. Override model selection, API keys, and parameters without changing core logic.

AI Content Repurposer: AI Recipe

What You Get

-Content analysis and key insight extraction
-Multi-format generation (tweets, LinkedIn, newsletter, blog)
-Platform-specific tone and length optimization
-Hashtag and keyword suggestion
-Content calendar with posting schedule

Step by Step

1. Ingest and analyze the content

Accept content in any format: blog post URL (fetch and extract text), YouTube transcript (via API), podcast transcription text, or raw text. Use OpenAI to analyze the content and extract the main thesis and structure.

2. Extract key insights

Use OpenAI to extract 5-7 key insights: quotable lines, surprising statistics, contrarian opinions, actionable takeaways, and the core narrative arc. Each insight includes a suggested format type (tweet-friendly, LinkedIn story, newsletter hook).

3. Generate format-specific versions

For each target format, generate optimized content: Twitter thread (5-8 tweets with hook and CTA), LinkedIn post (professional tone, 1200-2000 chars), newsletter blurb (150 words with subject line), blog summary (300 words with headings), and 3 Instagram caption options.

4. Optimize for each platform

Add platform-specific elements: relevant hashtags and keywords per platform, optimal posting times, character count checks, and tone adjustments (professional for LinkedIn, conversational for Twitter, visual for Instagram).

5. Build the content calendar

Generate a posting schedule: which format to post when, with suggested times for optimal reach. Include an A/B testing tracker to measure which repurposed formats get the best engagement.

Stack

OpenAINext.jsPostgreSQLResend API

Build This

Copy this prompt and paste it into Claude Code, OpenCode, Codex, or Cursor to build this recipe.

Build an AI content repurposer that transforms one piece of content into multiple platform-optimized formats. ROLE: You are a content transformation system that analyzes source content, extracts key insights, and generates platform-optimized versions for multiple distribution channels. CONSTRAINTS: - Content extraction must identify 5-7 key insights, quotes, or statistics per source piece - Each platform version must adhere to character limits: Twitter (280), LinkedIn (3000), Instagram (2200) - Tone must be adapted per platform: professional for LinkedIn, conversational for Twitter, visual for Instagram - Hashtag suggestions must be platform-specific: 3-5 for Twitter, 3-5 for LinkedIn, 5-10 for Instagram - Content calendar must respect optimal posting times per platform and timezone TOOL CALLING: - Use function calling for: analyze_content(content_url?, content_text?), extract_insights(content_id), generate_twitter(insights[], tone?), generate_linkedin(insights[], tone?), generate_instagram(insights[], tone?), generate_newsletter(insights[], word_count?), schedule_content(formats[], start_date?) - Each tool returns structured JSON with generated content and metadata STRUCTURED OUTPUT: - Content analysis must return JSON: { content_id: string, source_type: 'blog' | 'youtube' | 'podcast' | 'text', title: string, word_count: number, key_topics: string[], main_thesis: string } - Insights must return JSON: { insights: [{ id: string, type: 'quote' | 'statistic' | 'insight' | 'takeaway', content: string, suggested_format: string[] }], total_insights: number } - Platform content must return JSON: { platform: string, content: string, character_count: number, hashtags: string[], suggested_post_time: string, media_suggestions?: string[] } - Content calendar must return JSON: { posts: [{ platform: string, content: string, scheduled_time: string, format: string }], total_posts: number, duration_days: number } CHAIN OF THOUGHT: - Analysis: ingest content → identify format → extract structure → determine main themes - Insight extraction: scan for quotable lines → identify statistics → find actionable takeaways → assess relevance - Platform generation: adapt tone → adjust length → add platform-specific elements → optimize for engagement - Scheduling: determine optimal posting times → create calendar → balance platforms over time FEW-SHOT EXAMPLES: Source: Blog post 'How AI Agents Are Transforming Customer Service' Insights: [{ type: 'statistic', content: 'Companies using AI agents see 65% reduction in response time', suggested_format: ['twitter', 'linkedin'] }, { type: 'quote', content: '"AI agents don’t replace humans—they supercharge them"', suggested_format: ['twitter', 'instagram'] }] Twitter: { content: 'Thread: How AI agents are transforming customer service 🧵\n\n1/ Companies using AI agents see 65% reduction in response time. But here\'s what most people miss...', character_count: 247, hashtags: ['#AIAgents', '#CustomerService', '#Automation'] } Calendar: { posts: [{ platform: 'twitter', content: '...', scheduled_time: '2025-01-15T09:00:00Z', format: 'thread' }, { platform: 'linkedin', content: '...', scheduled_time: '2025-01-15T14:00:00Z', format: 'post' }], total_posts: 6, duration_days: 3 } EVALUATION CRITERIA: - Insight relevance: percentage of extracted insights that are meaningful and shareable - Platform optimization: percentage of posts that adhere to platform-specific best practices - Tone consistency: alignment between source content tone and platform-adapted tone - Calendar effectiveness: distribution of posts across optimal posting windows The system should: 1) Accept content in any format: blog post URL, YouTube transcript, podcast audio transcription text, or raw text, 2) Analyze the content and extract 5-7 key insights, quotes, and statistics, 3) Generate platform-optimized versions: a Twitter/X thread (5-8 tweets), a LinkedIn post with professional tone, a newsletter blurb (150 words), a blog summary (300 words), and 3 Instagram caption options, 4) Add relevant hashtags and keywords per platform, 5) Output a content calendar suggesting when to post each format for optimal reach, 6) Include an A/B testing tracker to measure which repurposed formats perform best.

Common Failure Modes

!Key insights miss the original content's main point
!Platform-specific optimization feels generic
!Tone doesn't match the original creator's voice
!Hashtag suggestions are irrelevant

Implementation Notes

For the first use, ask the user to provide 3 examples of their preferred writing style. Use these as few-shot examples for tone matching.

Ship ai content repurposer in production with 4M Labs

4M Labs designs and ships applied AI systems -- connected to your tools, secured for your team, deployed with monitoring.

Connected to your tools and data sources
Secured for your team with proper access controls
Deployed with monitoring and error handling
Documented for handoff and future maintenance

Work With 4M Labs

Frequently Asked Questions

Can I use this recipe in production?: Yes. Every recipe is production-tested with error handling, logging, and deployment guidance.
Which LLM providers are supported?: Recipes support OpenAI, Anthropic Claude, Google Gemini, and open-source models via a unified interface.
How do I customize these recipes?: Each recipe includes a configuration section. Override model selection, API keys, and parameters without changing core logic.