Objective-Driven Content Generator – OpenAI + Exa + Stability

This sequence is designed to efficiently achieve your desired objective by gathering all the necessary assets (text, images, and/or research) and generating a final output. Here’s a step-by-step overview of how it operates:

1. Defining the Objective

• The sequence starts by asking you:
  “What objective would you like the agent to achieve?”
• This helps establish the goal for the entire process.

2. Clarifying the Objective

• The system follows up with three additional questions to clarify and refine your objective.
• These questions ensure that the agent fully understands your requirements before proceeding.

3. The Loop: Gathering Assets

• After defining the objective, the system enters a loop to collect the assets needed to achieve it. The loop can run up to 20 times and performs one of the following actions in each iteration:
  • Draw an Image: Create visuals if they are needed for the objective. (Try using the phrase image or illustrated in your objective.)
  • Perform a Web Search: Gather information from the internet. (Include the phrase research or web search in your objective.)
  • Write Text: Generate content that fulfills parts of the objective.
• At the end of each loop, the system evaluates whether all required assets have been collected:
  • If yes, the loop ends early.
  • If no, the loop continues to gather missing assets.

4. The Final Phase: Creating the Output

• Once all assets have been gathered, the sequence moves to the final phase.
• Here, the system uses a text-to-text model to create a polished HTML-formatted output that aligns with your objective.