Prompt Templates for Common Tasks

The previous posts in this path covered prompt engineering as clear thinking and better prompting techniques. This post gives you something immediately practical: prompt templates — reusable frameworks for the tasks you do most often, designed to produce consistently useful results.

A prompt template is not a magic formula. It is a structured starting point that ensures you include the context, constraints, and specifications that produce good output. Over time, you will customize these templates to match your specific needs. The templates here give you a running start.

Writing Templates

Writing tasks are the most common AI use case, and the difference between vague and specific prompts is dramatic.

Email drafting. Specify the recipient, purpose, tone, and any constraints. "Write a professional email to [recipient/role] regarding [topic]. The purpose is to [request/inform/follow up]. Tone should be [formal/friendly/direct]. Keep it under [length]. Key points to include: [bullet points]. Do not include: [anything to avoid]."

The specificity matters. "Write an email about the project" produces generic filler. "Write a concise email to my product manager explaining that the API integration will be delayed by one week because the vendor changed their authentication method. Tone: direct but collaborative. Suggest two mitigation options." produces something you can send with minor edits.

Content creation. Start with the audience and purpose. "Write a [format: blog post/article/social post] for [audience] about [topic]. The purpose is to [educate/persuade/entertain]. Length: approximately [word count]. Tone: [specify]. Key points to cover: [list]. Do not include: [things to avoid]. End with: [call to action or summary style]."

Editing and revision. When you have existing text that needs improvement. "Review the following text and [improve clarity/fix grammar/make more concise/adjust tone to be more professional]. Preserve the original meaning and key points. Highlight any factual claims that should be verified. Here is the text: [paste text]."

Analysis Templates

Analysis tasks — summarizing, comparing, evaluating — benefit from clear instructions about depth, format, and focus.

Document summary. "Summarize the following document in [number] paragraphs. Focus on [specific aspects or themes]. The audience is [who will read this summary]. Include: [specific elements to capture]. Exclude: [details to omit]. Note any conclusions or recommendations the document makes. Here is the document: [paste or describe]."

Comparison. "Compare [option A] and [option B] across these dimensions: [list dimensions]. For each dimension, explain the strengths and weaknesses of each option. Conclude with a recommendation for [specific context or use case]. Present the comparison in [format: prose/table/structured sections]."

Data interpretation. "Analyze the following data and identify [trends/patterns/anomalies/insights]. Context: this data represents [what the data measures] over [time period] for [entity]. Key questions to answer: [specific questions]. Note any caveats about data quality or limitations of the analysis. Here is the data: [paste data]."

Coding Templates

Coding prompts need precision about language, constraints, and the surrounding context.

Code generation. "Write a [language] function that [specific behavior]. Input: [describe inputs with types]. Output: [describe expected output with type]. Edge cases to handle: [list]. Include error handling for [specific error scenarios]. Follow [coding convention or style guide]. Include comments explaining the approach."

Code review. "Review the following [language] code for [bugs/performance issues/security vulnerabilities/readability]. Explain any issues you find, why they are problems, and how to fix them. Rate the overall code quality and suggest the most impactful improvements. Here is the code: [paste code]."

Debugging. "I am getting [error message or unexpected behavior] when running this [language] code. Expected behavior: [what should happen]. Actual behavior: [what happens instead]. Environment: [relevant details — OS, runtime version, dependencies]. I have already tried: [what you have attempted]. Here is the relevant code: [paste code]."

The debugging template is particularly important. AI assistants are dramatically more helpful when they understand what you expected, what actually happened, and what you have already tried — the same information that makes Stack Overflow questions useful.

Decision-Making Templates

When you need to think through a decision rather than produce content.

Pros and cons analysis. "I am deciding whether to [decision]. Context: [relevant background]. Help me think through this by listing pros, cons, and risks. For each pro and con, rate the importance (high/medium/low) and explain your reasoning. Flag any information I should gather before deciding."

Technical decision. "I need to choose between [option A] and [option B] for [specific use case]. Requirements: [list must-haves]. Nice-to-haves: [list preferences]. Constraints: [budget/timeline/team skills/existing infrastructure]. Evaluate each option against these criteria. Recommend one and explain the reasoning, including what trade-offs the recommendation involves."

Problem decomposition. "I am facing this problem: [describe problem]. Help me break it into smaller, manageable sub-problems. For each sub-problem, suggest an approach and estimate the relative difficulty. Identify which sub-problems should be tackled first and which depend on others."

Customization Workflow

These templates are starting points. The customization process:

Start with the template that matches your task. Fill in the specifics. Submit it and evaluate the result. Note what was good and what needed improvement. Modify the template based on what you learned — add constraints that the first result missed, remove instructions that produced unwanted output, adjust the tone or format.

After a few iterations, you will have personalized templates that consistently produce good results for your specific context. Save these. A "weekly report prompt" that includes your team's format, your manager's preferences, and the data sources you use every week is a tool that saves time every time you use it.

The meta-prompt for building templates: "I frequently need to [task]. My audience is [who]. My preferences are [style/format/tone]. Help me create a reusable prompt template that I can fill in each time to get consistently good results."

The Takeaway

Prompt templates eliminate the cold-start problem of writing prompts from scratch. They ensure you include the context, constraints, and specifications that produce useful output. Start with these general frameworks, customize them through iteration, and build a personal library of templates for your recurring tasks.

The template is not the goal — the consistently useful output is. The template just makes getting there faster and more reliable.

---

Next in the "Prompting and Working with AI" learning path: We'll cover working with AI on complex projects — how to use AI as a thought partner for research, planning, and iterative problem-solving rather than just one-shot tasks.