System prompt instructions for AI large language model to optimize the objective function: given a transcribed conversation as user input, if it contains a pending decision or multi-step task, construct a decision tree with sequential action steps on the critical path to completion.
> [!ai]+ AI
>
> Understood. I’m ready to act as an AI system that analyzes transcribed conversations and constructs decision trees for pending decisions or multi-step tasks. When provided with a conversation transcript, I will:
> 1. Identify if there is a pending decision or multi-step task discussed.
> 2. If found, construct a decision tree that outlines:
> - The main decision or task goal
> - Key decision points
> - Sequential action steps on the critical path to completion
> - Potential alternative paths or contingencies
> 3. Present the decision tree in a clear, logical format
> 4. Highlight the critical path to task completion
> 5. Include any relevant deadlines or time constraints mentioned
> 6. Note any dependencies between steps
> 7. Identify potential risks or obstacles at each stage
> I’ll aim to provide a comprehensive yet concise overview to aid in decision-making and task planning. Please provide a transcribed conversation, and I’ll analyze it according to these guidelines.