prompt-factory

Generate production-ready LLM prompts via a 5-7 question flow and 69 role presets.

1|Updated May 6, 2026
One-click install
npx skills add https://github.com/surfingalien/FinSurfing --skill prompt-factory-surfingalien
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prompt-factory
Source: https://github.com/surfingalien/FinSurfing/tree/main/.claude/skills/prompt-factory
Command: npx skills add https://github.com/surfingalien/FinSurfing --skill prompt-factory-surfingalien

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Writing effective prompts for LLMs like Claude, ChatGPT, and Gemini usually requires many iterations and deep prompt-engineering knowledge. This Skill turns a rough idea into a validated, ready-to-use mega-prompt in one pass. ## Core Features & Use Cases - Guided Question Flow: Asks 5-7 mandatory questions about role, domain, constraints, and style to capture the context a good prompt needs. - 69 Quick-Start Presets: Covers 15 professional domains including engineering, product, marketing, legal, finance, HR, and C-suite roles. - Multi-Format Output: Generates prompts in XML, Claude, ChatGPT, or Gemini formats, in Core (~5K tokens) or Advanced (~10K tokens) mode, with 7-point quality validation. - Use Case: A product manager says "Use the Senior Product Manager preset" and receives a complete XML mega-prompt with workflow, best practices, and examples, ready to paste into any LLM. ## Quick Start Ask the assistant to use the Prompt Factory to create a prompt for your role, or name one of the 69 presets to load it directly.

Frequently Asked Questions about prompt-factory

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

FAQPage Schema
How do I create a prompt for a specific professional role?

Request a preset by name, such as "Use the Senior Full-Stack Engineer preset", or describe a custom role and answer 5-7 questions about domain, task, constraints, and style. The Skill then generates a validated mega-prompt in your chosen format.

What output formats does the prompt generator support?

It supports four formats: XML (default, structured tags optimal for LLM parsing), Claude system prompt format, ChatGPT custom instructions format, and Gemini format. You can also request all four formats at once for team distribution.

What is the difference between Core and Advanced mode?

Core mode delivers the prompt plus usage instructions and 2-3 examples at roughly 4-6K tokens. Advanced mode adds 5 testing scenarios, 3 prompt variations (concise, balanced, comprehensive), and optimization tips at roughly 8-12K tokens.

Does this Skill implement the work described in the generated prompt?

No. It only generates the prompt document and stops after delivery. You copy the prompt into a fresh conversation or another tool, where the actual implementation work happens.

How can I generate prompts for an entire team at once?

Use the batch_generator.py script with a CSV or JSON file listing role, domain, goal, and tech stack per prompt. It generates all prompts in parallel into an output directory and writes a summary report with validation results.

Why does the Skill always ask questions even for obvious requests?

The mandatory 5-7 question flow validates assumptions about domain, constraints, and success criteria that users often omit. Skipping questions historically produced generic prompts, so the flow enforces minimum context gathering before generation.