PromptIQ: An AI-Assisted Prompt Analysis and Refinement System Using Chrome Extension, JavaScript and Rule-Based Scoring

Abstract

Large language models such as ChatGPT, Gemini and Claude are increasingly used for education, software development, data analysis, content creation and decision support. The quality and usefulness of their responses are strongly influenced by how users formulate prompts. However, many users do not have a systematic way to judge whether a prompt is clear, contextual, specific and efficient before submitting it. This paper presents PromptIQ, a browser-based prompt analysis and refinement system designed to provide immediate feedback while a user is interacting with supported AI platforms. The proposed system uses a Chrome extension to detect prompt text through a content script and pass it to a local JavaScript scoring engine. The current scoring model evaluates four dimensions: clarity, context, specificity and efficiency, combined into an overall score on a 0-100 scale. Functional testing of the implemented modules is used to evaluate the reliability of the system. The proposed system demonstrates how browser-based and rule-based technologies can be combined to develop an accessible, transparent prompt-quality feedback platform, with a future version expected to integrate an external language model for explanation and automatic prompt refinement. Keywords— Prompt Engineering, Large Language Models, Prompt Analysis, Prompt Optimization, Chrome Extension, Natural Language Processing, Generative AI.