As marketers and content creators, we all know the struggle of repurposing content for different platforms. Creating fresh content is hard enough, but adapting it for various channels can be an even bigger headache. In this guide, we’ll show you how to automate your content repurposing workflow by building a custom AI Agent with Tars.

The Problem with Content Repurposing

Content creation is time-consuming, but the real challenge often comes when you need to transform that same content for different platforms:

• Converting a blog post into a LinkedIn article
• Turning a video script into a Twitter thread
• Adapting website content for an email newsletter

Each platform requires different formatting, length, and style – making the repurposing process tedious and time-consuming.

Introducing the Tars Content Repurposing AI Agent

Our AI agent will help you automatically repurpose any piece of content into multiple formats:

• Blog posts
• Social media posts (LinkedIn and Twitter)
• Email newsletters

The best part? You can fully customize it to match your specific content needs and brand voice.

Step-by-Step Guide to Building Your Agent

1. Creating the Welcome Message

First, let’s set up the initial interaction:

1. Click the “Add a Gambit” button (a Gambit is one step in the conversation)
2. Name it “welcome_message” (use lowercase with underscores for naming)
3. Edit the welcome message to introduce your AI agent
4. Create button options for content input methods:
   • Provide a URL
   • Copy and paste content
5. Enable branching and detach mode in settings

2. Setting Up Content Input Methods

Now we’ll create two paths based on how users want to input their content:

For URL Input:

1. Create a Gambit named “link_prompt”
2. Make it a no-input Gambit
3. Connect it to the URL button from the welcome message

For Copy/Paste Input:

1. Create a Gambit named “paste_prompt”
2. Make it a no-input Gambit
3. Connect it to the copy/paste button

3. Creating AI Agent Gambits for Content Analysis

For each input method, we need an AI agent to analyze the content:

For URL Input:

1. Create an AI Agent Gambit named “url_agent”
2. Edit the base prompt to instruct the agent to scrape the URL with FireCrawl
3. Clear default messages
4. Attach the FireCrawl tool

For Paste Input:

1. Create an AI Agent Gambit named “paste_agent”
2. Edit the base prompt to analyze pasted content
3. Clear default messages
4. Attach a tool (like JIGY search) even if not used

4. Setting Up Content Repurposing Options

After content analysis, we’ll offer repurposing options:

1. Create a Gambit named “user_choice”
2. Add buttons for different repurposing options:
   • Blog post
   • Social media posts
   • Email newsletter
3. Enable branching and detach mode

For social media specifically:

1. Create a sub-Gambit for platform selection
2. Add buttons for LinkedIn and Twitter
3. Enable branching and detach mode

5. Building the Blog Post Path

Let’s create a sequence to generate blog posts:

1. Keyword Input:
   • Create a Gambit named “blog_keyword”
   • Ask for target keyword
   • Use text input
2. Blog Length:
   • Create a Gambit named “blog_length”
   • Add button options (500 words, 500-750 words, 1000+ words)
   • Disable branching
3. Blog Structure:
   • Create a Gambit named “blog_sections”
   • Add button options for number of sections
   • Disable branching
4. Call-to-Action:
   • Create a Gambit named “blog_cta”
   • Use text input for custom CTA
5. Preference Summary:
   • Create a Gambit named “blog_user_preferences”
   • Summarize all user selections using {ursp_gambit_name} references
   • Make it a no-input Gambit
6. Blog Generation Agent:
   • Create an AI Agent Gambit named “agent_blog”
   • Create a detailed prompt referencing user preferences
   • Enable chat history
   • Attach a tool (even if not used)
   • Loop the node back to itself for iterations

6. Building the Social Media Paths

LinkedIn Post Path:

1. Writing Style:
   • Create a Gambit for LinkedIn writing style preferences
   • Use button options
   • Disable branching
2. Post Length:
   • Create a Gambit for LinkedIn post length
   • Use button options
   • Disable branching
3. Target Audience:
   • Create a Gambit for audience targeting
   • Use text input
4. Call-to-Action:
   • Create a Gambit for LinkedIn CTA
   • Use text input
5. Preference Summary:
   • Create a summary Gambit
   • Make it a no-input Gambit
6. LinkedIn Post Generation Agent:
   • Create an AI Agent Gambit with detailed instructions
   • Include examples of good LinkedIn posts
   • Enable chat history
   • Attach a tool
   • Loop back for iterations

Twitter Thread Path:

1. Thread Length:
   • Create a Gambit named “tweet_number”
   • Add button options (4-6 tweets, 7-9 tweets, 10-12 tweets)
   • Disable branching
2. Tweet Style:
   • Create a Gambit named “tweet_style”
   • Add style options (narrative, humorous, direct)
   • Disable branching
3. Additional Details:
   • Create a Gambit named “tweet_details”
   • Use text input for context, audience, etc.
4. Preference Summary:
   • Create a summary Gambit
   • Make it a no-input Gambit
5. Twitter Thread Generation Agent:
   • Create an AI Agent Gambit with detailed instructions
   • Include example threads
   • Enable chat history
   • Attach a tool
   • Loop back for iterations

7. Building the Email Newsletter Path

1. Newsletter Type:
   • Create a Gambit named “email_newsletter_type”
   • Add button options (promotional, educational, product update)
   • Disable branching
2. Brand Context:
   • Create a Gambit named “email_brand_context”
   • Use text input for brand information
3. Newsletter Length:
   • Create a Gambit named “email_newsletter_length”
   • Add button options (short, medium, long)
   • Disable branching
4. Call-to-Action:
   • Create a Gambit named “email_cta”
   • Use text input
5. Preference Summary:
   • Create a summary Gambit
   • Make it a no-input Gambit
6. Newsletter Generation Agent:
   • Create an AI Agent Gambit with detailed instructions
   • Enable chat history
   • Attach a tool
   • Loop back for iterations

8. Finalizing Your Agent

Once all paths are complete:

1. Go to the design section to customize the look and feel
2. Choose a theme
3. Name your agent
4. Save changes and publish

Key Tips for Effective AI Agents

• Naming Convention: Use lowercase with underscores for all Gambits
• AI Agent Prompts: The “base prompt” is where the magic happens – be specific with instructions
• Provide Examples: Give examples in your prompts for better outputs
• Tool Attachment: AI Agent Gambits need a tool attached (even if not used)
• Loop for Iterations: Connect AI Agent Gambits back to themselves for continued refinement

Conclusion

By building this AI agent, you can save countless hours on content repurposing. The best part is that you can fully customize it to match your specific content needs, brand voice, and marketing strategy. No matter what content you create, your TARS AI agent can help you maximize its reach across multiple platforms with minimal effort.

Ready to automate your content repurposing workflow? Try our pre-built content repurposer at hellotars.com/ai-agents/content-repurposer or get started building your own with Tars today!

For more tutorials and resources, visit hellotars.com/blog or join our community at hellotars.com/discord.