prompt-master

Generate production-ready prompts optimized for specific AI tools and models.

Updated Jul 1, 2026
One-click install
npx skills add https://github.com/MilUX-Ltd/footing --skill prompt-master-milux-ltd
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: prompt-master
Source: https://github.com/MilUX-Ltd/footing/tree/main/footing/template/Skills/meta-skills/prompt-master
Command: npx skills add https://github.com/MilUX-Ltd/footing --skill prompt-master-milux-ltd

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Writing prompts that work reliably across different AI tools is hard because each model and platform has distinct syntax, behavior, and failure modes. This Skill converts a rough idea into a single copy-paste-ready prompt tailored to the target tool, eliminating trial-and-error re-prompting. ## Core Features & Use Cases - Tool-Specific Prompt Generation: Produces prompts formatted for Claude, Claude Code, Cursor, Copilot, GPT models, reasoning models (o3, DeepSeek-R1), Gemini, browser agents, and research orchestrators. - Diagnostic Repair: Scans existing prompts for failure patterns like vague tasks, missing output formats, absent stop conditions, or Chain-of-Thought applied to reasoning-native models, then fixes them silently. - Prompt Decompiler: Takes a pasted prompt and rebuilds it for a different target tool while preserving intent and constraints. - Use Case: You need Cursor to refactor a function without touching anything else. Describe the change, and receive a prompt with file anchors, a do-not-touch list, and explicit done criteria ready to paste. ## Quick Start Ask the assistant to write a prompt for your target tool, for example: write a prompt for Claude Code that migrates my database schema with stop conditions before any destructive action.

Frequently Asked Questions about prompt-master

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write a prompt for Claude Code or Cursor?▼

Provide the file path, function name, current behavior, desired change, and a do-not-touch list. For agentic tools like Claude Code, also include starting state, target state, allowed actions, and explicit stop conditions before destructive operations.

Should I add Chain of Thought to prompts for o3 or DeepSeek-R1?▼

No. Reasoning-native models like o3, o4-mini, DeepSeek-R1, and Qwen3-thinking reason internally, and adding Chain of Thought or step-by-step scaffolding degrades their output. Use short, clean zero-shot instructions instead.

Can I adapt an existing prompt for a different AI tool?▼

Yes. Paste the existing prompt and name the new target tool. The decompiler mode extracts the original intent and constraints, then rebuilds the prompt using the syntax and conventions of the new tool.

Why does my prompt produce inconsistent output formats?▼

Inconsistent formats usually result from missing explicit format locks. Add a labelled output example, specify exact structure and length, and state what done looks like. Models like Gemini especially need explicit format constraints to avoid drift.

What are the limitations of single-prompt techniques?▼

Techniques like Mixture of Experts, Tree of Thought, and Graph of Thought embedded in one prompt are fabrication-prone and unreliable. Complex tasks should be split into sequential prompts with verification checkpoints instead.