How does the comparison feature work?

It takes two Prompt IDs and analyzes their metadata and elements to calculate a similarity score, listing shared components and unique elements for both.

What is required for this skill to work?

It relies on a database (elements.db) containing historical prompt data, typically generated by an associated prompt generation skill.

Does this skill generate new prompts?

No, the Prompt Analyzer is dedicated exclusively to analyzing, comparing, and providing insights into prompts that have already been generated and saved in your database.

What kind of statistics does the element library provide?

It provides usage frequency, average quality scores per element, and distribution across categories like makeup styles, lighting, and clothing.

Can it suggest elements for a specific style?

Yes, it can recommend the best combinations of elements for styles like 'Cyberpunk' or 'Ancient Chinese' based on historical success and usage frequency.

Prompt Analyzer & Insights

Name: Prompt Analyzer & Insights
Author: tianzecn

bytianzecn

•

Data Science & ML

Analyzes generated prompts to provide deep insights into element usage, quality comparisons, and style-based recommendations.

Prompt Analyzer is a specialized utility for Claude Code designed to optimize prompt engineering workflows. It focuses on post-generation analysis, allowing users to inspect the specific elements that comprise a prompt, compare two different prompts for similarity, and query element library statistics. By leveraging historical data, it can recommend similar prompts and suggest optimal element combinations for specific styles such as Sci-Fi, Cyberpunk, or Ancient Chinese. This skill is essential for users looking to audit their prompting history, understand why certain outputs succeed, and build a high-quality library of reusable prompt components.

Key Features

01Side-by-side prompt comparison with similarity scoring and unique attribute analysis

02Detailed element inspection including quality scoring and style tags

03Similarity-based prompt recommendations derived from historical data

04Style-based element grouping for optimized creative direction

05Comprehensive element library statistics for usage frequency and performance

061 GitHub stars

Use Cases

01Auditing historical prompt performance to identify high-quality element patterns

02Debugging output variations by comparing the structural differences between two prompt versions

03Finding the most effective element combinations for specific visual or narrative styles

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add tianzecn/myclaudecode prompt-analyzer

For use in Claude.ai and ChatGPT

Download Skill