I love a good prompt—especially the ones that do more than crank out limericks like a caffeinated leprechaun. Don’t get me wrong, I appreciate a limerick as much as the next person who’s given up on dignity.
What I truly enjoy are the prompts that drag me, begrudgingly, up to a 10,000-foot view of my existence so I remember what I’m actually supposed to be doing before I vanish in a puff of poor decisions. So I figured I’d start sharing my favorite AI Prompts here from time to time, like a survival kit for people whose minds wander off and don’t leave a forwarding address – it’s ok, my brain needs both alliteration and supervision too.
The following prompt was handed to me by Microsoft MVP Pierre-Yves Delacote—easily one of the most interesting human beings I’ve had the fortune of collaborating with across dozens of Microsoft Copilot and AI projects over the last couple of years. Seriously, working with him is like opening a mystery box that always contains brilliance, mild chaos, and the unsettling realization that he thought ten steps ahead while I was still looking for coffee.
Token Trail: preparing to pop smoke for pto or a holiday
With the ‘Work’ tab selected in Microsoft Copilot Chat, paste the following prompt and send it:
“Act as my productivity assistant. I’m preparing to go on holiday from 11/21 to 11/28. Based on my role, responsibilities, current projects, and collaborators, help me create a clear, prioritized list of everything I should complete or delegate before I log off — to ensure a smooth transition and peace of mind. Please organize the output into the following sections: Critical Work Tasks – Review my calendar, flagged emails, and open tasks from the last 2 weeks only to identify must-do items before I leave. Present this section in a table with the following columns: Task Due Date Owner Notes/Dependencies Reference (Source) Reference Link Handover & Delegation Tasks – Identify ongoing work that can be delegated based on recent emails, chats, and shared task ownership from the last 2 weeks only. Present this section in a table with the following columns: | Task | Backup Owner | Access Needed | Status/Instructions | Reference (Source) | Reference Link | Communication & Planning Tasks – Draft OOO messages and stakeholder notifications based on my calendar and recent communications. Team & Leadership Alignment – Summarize key updates for my manager, skip manager, and team based on recent communications and shared documents. Personal Wrap-Up – Suggest wrap-up steps based on my current workload and calendar to help me disconnect stress-free. Use my Microsoft 365 data (Outlook, Teams, Planner, Loop, To Do) from the last 2 weeks only to tailor recommendations. Include direct links to the original email, chat, or document where each task was sourced (when available). End with a concise summary of the Top 5 tasks to finish before I go, formatted as a checklist I can easily add to my calendar.”
Tangent: Why words matter
Whether you’re using AI for creative writing, coding assistance, data analysis, or problem-solving, the quality of your input directly shapes the quality of your output – garbage in, garbage out. Think of it as learning a new language—not of code or foreign grammar, but of clear, intentional communication with machines that process language in ways both familiar and alien to our own thinking.
I’ll admit, when I first started working with GenAI, I treated the models like search engines. But AI models, no matter how sophisticated, operate on the principle of pattern matching and contextual understanding. They don’t truly “know” what you want—they interpret your request based on the patterns they’ve learned. A vague prompt like “write something good” leaves enormous room for misinterpretation (although it would be pretty fricken funny if the response was “Something good.”), while a specific prompt like “write a 300-word professional email declining a meeting invitation while suggesting alternative times” gives the model clear parameters to work within.
Unlike human conversation partners who share your physical context and can pick up on non-verbal cues, GenAI only has the text you provide. Every piece of context you include—your audience, your goals, your constraints—helps the model better understand not just what you want, but why you want it and infer what else you need. This understanding leads to outputs that are not just accurate, but actually useful.
prompt construction
Well-crafted initial prompts reduce the need for endless back-and-forth refinements. When you front-load your prompt with relevant details, specifications, and examples, you’re more likely to get usable results on the first try. This saves time and reduces frustration, transforming AI from a tool that requires constant tweaking into one that delivers reliable results.
Before diving into examples, let’s understand what makes a prompt effective:
Purpose: Explain how the output will be used
Clarity: Use specific language rather than ambiguous terms
Context: Provide background information relevant to your request
Constraints: Specify any limitations (length, format, style, etc.)
Examples: When possible, show what you’re looking for
prompt Style: ThAT strategic Analyzer PERSON
Why it works: This prompt provides specific parameters while maintaining flexibility for comprehensive analysis. It guides structure without being overly prescriptive.
“Analyze the pros and cons of [specific decision] for a [type of organization] with approximately [size] employees. Consider factors including cost, implementation time, employee impact, and long-term scalability. Present your analysis in a structured format with clear sections.”
prompt Style: The CODE DEBUGGER THINGY
Why it works: It asks for multiple layers of analysis (bugs, security, performance) and requires both identification and education, making the response more valuable for learning.
“Review this code for potential bugs, security vulnerabilities, and performance issues: [code snippet]. Explain each issue you find, why it’s problematic, and provide a corrected version with comments explaining the changes.”
