Content Blueprint AI 💡

Generates focused, high-quality creative content blueprints (like scripts or outlines) for a single selected platform (e.g., TikTok, Blog, YouTube) based on analyzed trend data.

How it Works

This actor is designed as the final stage in the AI Content Autopilot pipeline. It takes the dataset produced by the Sentiment Compass actor as input. This dataset contains rich information about emerging trends, including sentiment analysis and key entities.

For each topic found in the input dataset, this actor:

1. Analyzes the topic data (sentiment, tone, entities).
2. Makes a single, targeted call to an AI language model (like GPT-4o) using a specialized prompt designed for the specific content format you selected.
3. Generates a detailed, structured blueprint for that format.
4. Outputs the original topic data, augmented with the newly generated blueprint.

Why only one format per run?

• Quality Focus: By concentrating the AI's processing power (token budget) on a single format, it produces a more detailed and higher-quality blueprint compared to generating multiple formats simultaneously.
• Cost Management: Each run makes only one LLM call per topic for the selected format, making API usage predictable and cost-effective. The necessary OpenAI API key is pre-configured by the actor owner via secure environment variables, so you do not need to provide your own key.

Input Configuration

Configure the actor using the following input fields:

1. Source Dataset (dataset_id): Required. Select the dataset generated by the Sentiment Compass actor. This dataset contains the analyzed topics that the blueprints will be based on. Use the dataset picker to find the correct one.
2. Output Content Format (output_format): Required. Choose the single content format you want to generate blueprints for in this run. Options include:
   • Blog
   • TikTok
   • YouTube
   • AdCopy
   • Instagram Post
   • LinkedIn Post
   • X (Twitter) Thread
   • Newsletter
   • Podcast Script
3. Max Topics to Process (max_topics_to_process): Optional (Default: 10). Limits how many topics from the source dataset will be processed in this run. Helps manage run time and cost for large datasets.
4. Creativity Level (creativity_level): Optional (Default: 70%). Controls the AI's creativity (temperature). Higher values (e.g., 80-100) produce more novel ideas, while lower values (e.g., 20-50) result in more focused and predictable output.

Output Structure

The actor outputs the original data for each topic received from the Sentiment Compass, adding a new field called generated_blueprints.

• generated_blueprints: An array containing a single object. This object has two keys:
  • format: The content format you selected (e.g., "TikTok").
  • blueprint: A JSON object containing the detailed, structured content blueprint generated by the AI for that format. If the generation fails (e.g., due to an API error), this field will contain an error message.

Example Output Item Snippet:

{
  "cluster_topic": "AI in Cybersecurity",
  "average_sentiment": "Negative",
  "sentiment_score": -0.65,
  "emotion_profile": { "fear": 0.8, "crisis": 0.7, ... },
  "volatility_score": 75,
  "recommended_tone": "Cautionary and informative",
  "entity_sentiments": [ { "name": "Google", "sentiment": "Neutral" }, ... ],
  "source_cluster_id": "TREND_AI_IN_CYBERSECURITY_...",
  // --- Added by this actor ---
  "generated_blueprints": [
    {
      "format": "Blog",
      "blueprint": {
        "title": "The Double-Edged Sword: AI's Role in Modern Cybersecurity Threats",
        "introduction": "...",
        "section_1_title": "AI as a Defender...",
        "section_1_content": "...",
        "section_2_title": "AI Empowering Attackers...",
        "section_2_content": "...",
        "section_3_title": "Staying Ahead of AI Threats...",
        "section_3_content": "...",
        "conclusion": "..."
      }
    }
  ]
}

## Integration with AI Content Autopilot

When using the main AI Content Autopilot orchestrator:

    Ensure the "Run Content Idea Generator" stage is enabled.

    Select the desired content format(s) in the orchestrator's input settings under "Output Content Formats".

    The orchestrator will automatically call this Content Blueprint AI actor, passing the correct dataset ID from the previous stage and the selected format.