Prompt Patterns III: Refining and Expanding

Hi Rob! 😊 No worries—your Prompt Patterns III content is safe in my memory bank, and I’m ready to create it for you whenever you like. Let’s dive into it and complete the trilogy, shall we?

Prompt Patterns III: Refining and Expanding

As the course reached its final module, the focus shifted toward fine-tuning outputs and combining techniques to achieve greater precision and adaptability. These patterns act as the ultimate toolkit for transforming basic prompts into powerful, multi-layered solutions.

1. Ask for Input Pattern: Collaborate with AI

This pattern invites AI to become an active participant in problem-solving, much like brainstorming with a teammate. By asking for input or clarification, you can refine the AI’s responses for better outcomes.

• Example Prompt:
  “I’m drafting an email to introduce a new feature to our customers. What tone do you think would work best: formal, friendly, or professional? Why?”
• Result: The AI not only suggests a tone but explains its reasoning, allowing you to choose or refine further.

2. Combining Patterns: Power in Unity

Just as agile teams thrive by integrating diverse skills, combining patterns allows the AI to handle complex tasks.

• Example:
  Combine the Chain-of-Thought and ReAct patterns to solve a problem step-by-step and take action.
  Prompt:
  “You are an agile coach analyzing a failed sprint. Think step-by-step about what went wrong, then suggest three actions to improve the next sprint.”
• Result: The AI identifies root causes and provides actionable solutions, mimicking a full retrospective.

3. Outline Expansion Pattern: From Skeleton to Full Body

This pattern takes an initial outline or idea and expands it into a detailed response. It’s perfect for breaking down complex ideas or generating content.

• Story: I once asked the AI to help structure a workshop. Using the outline expansion pattern, I provided a skeleton of topics, and it fleshed them out beautifully.
  Prompt:
  *“Here’s a rough outline for a Scrum training:
  1. Introduction to Scrum.
  2. Key Roles.
  3. Benefits of Scrum.
     Expand this into a detailed agenda with time blocks.”*
• Result: The AI produced a full agenda, complete with discussion points and estimated durations.

4. Menu Actions Pattern: Offer Choices

This pattern mirrors the decision-making process by presenting a menu of options. It’s great for scenarios requiring user input or multiple paths.

• Example Prompt:
  *“We’re prioritizing the next sprint backlog. Here are three options:
  1. Fix high-priority bugs.
  2. Develop a new feature.
  3. Improve existing documentation.
     What do you recommend, and why?”*
• Result: The AI provides a rationale for each option, helping you make an informed decision.

5. Fact Check List Pattern: Validate and Verify

This pattern ensures accuracy by having the AI cross-check information against known facts. It’s invaluable for agile documentation or presentations.

• Example Prompt:
  “Here’s a claim: ‘Scrum teams should have 15+ members.’ Is this correct? Provide supporting evidence.”
• Result: The AI identifies the inaccuracy and explains that Scrum teams typically have 3–9 members, aligning with the Scrum Guide.

6. Tail Generation Pattern: Keeping the Conversation Going

Tail generation ensures continuity, keeping the conversation or task on track. It’s ideal for follow-ups in agile workflows.

• Example Prompt:
  *“Here’s our retrospective summary:
  1. Improve communication in stand-ups.
  2. Reduce technical debt in the next sprint.
     What should our next steps be?”*
• Result: The AI suggests actionable follow-ups, maintaining momentum toward improvement.

7. Semantic Filter Pattern: Narrowing the Focus

This pattern helps refine responses by filtering irrelevant information, ensuring the output stays on-topic.

• Example Prompt:
  “Focus only on issues related to team communication in this retrospective feedback: [Insert Feedback].”
• Result: The AI filters out unrelated comments and hones in on communication-related insights, making retrospectives more efficient.

Conclusion for Prompt Patterns III

This final module ties everything together, showing how to refine prompts and combine techniques for maximum impact. Whether it’s asking for input, validating facts, or expanding ideas, these patterns demonstrate the flexibility and power of AI as a collaborative tool.