Generate and Refine Code Prompts

This skill is a library of ready-to-use prompt templates, described as battle-tested and inspired by the awesome-chatgpt-prompts collection and community best practices. It gives an assistant a catalog of prompts it can hand to a user or apply directly, grouped by purpose, along with prompt engineering techniques and a checklist for improving prompts.

The prompt categories are:

• Role-based prompts: Expert Developer (15+ years of experience, clean code, SOLID principles, pragmatic architecture), Code Reviewer (bugs, edge cases, naming, security, with a severity-coded review format), Technical Writer (clear, jargon-free documentation), and System Architect (scalability, reliability, maintainability, performance, cost, with trade-off analysis).
• Task-specific prompts: Debug This Code, Explain Like I'm 5, Code Refactoring, Write Tests, and API Documentation.
• Analysis prompts: Code Complexity Analysis, Performance Analysis, and Security Review, with issues classified by severity from Critical to Low.
• Creative prompts: Brainstorm Features (ten ideas, with the top three ranked by impact and effort) and Name Generator (ten options across descriptive, evocative, acronym, and metaphorical categories).
• Transformation prompts: Migrate Code between languages and Convert Format between data formats.

The prompt engineering techniques are Chain of Thought, Few-Shot Learning, the Persona Pattern, and Structured Output using a JSON response format with analysis, recommendations, confidence, and caveats.

Use it when the user needs ready-to-use prompt templates, wants role-based prompts of the "act as X" kind, asks for prompt examples or inspiration, needs task-specific prompt patterns, or wants to improve their prompting.

Do not expect it to run tasks on its own: it supplies wording, and templates with placeholders such as a concept, framework, product, or source and target formats must be filled in by the user. The skill is marked critical risk in its metadata, and its limits state that output is not a substitute for environment-specific validation, testing, or expert review, and that the assistant should stop and ask for clarification if inputs, permissions, safety boundaries, or success criteria are missing.

Input is the user's goal and the values for the template placeholders. Output is a prompt, or a response shaped by one. For example, the Chain of Thought template reads:

Let's solve this step by step:
1. First, I'll understand the problem
2. Then, I'll identify the key components
3. Next, I'll work through the logic
4. Finally, I'll verify the solution
[Your question here]

The Debug This Code template asks for problem identification, root cause, a fix with corrected code, and prevention, showing the debugging thought process step by step. The Write Tests template asks for happy paths, edge cases, error conditions, and boundary values, using descriptive names and the Arrange-Act-Assert pattern.

The skill has no runtime integrations. It points to outside reading: the awesome-chatgpt-prompts repository, the prompts.chat site, and the Learn Prompting site. Prompts can be pasted into any assistant.

The prompt improvement checklist asks whether you have a clear objective, provided context, specified the output format, given examples, defined constraints, and stated success criteria. The closing tip is that the best prompts are specific, provide context, and include examples of the desired output.

It is for developers, writers, and analysts who want reliable starting points for everyday prompting tasks such as code review, debugging, testing, documentation, brainstorming, and migration, and for anyone learning to write better prompts.