Inspiration We built Prompt Pal because many people know what they want help with, but struggle to explain it clearly to an AI tool. This is especially true for students working with assignment instructions, screenshots, files, rubrics, and course pages. A weak prompt often leads to a weak answer, so we wanted to create a tool that helps users turn messy first thoughts into focused, useful prompts.

What it does The project is a Chrome extension that helps users turn rough ideas into better AI prompts.

It lets users enter a basic prompt, attach files or screenshots, add academic or D2L Brightspace course context, and choose settings like tone or resume/career mode. Then it generates a clearer, more detailed prompt that can be copied or sent into AI tools like ChatGPT, Claude, Gemini, Perplexity, or Copilot.

It also supports AI-powered prompt improvement using API keys for providers like OpenAI, Claude, Gemini, and Perplexity, while still offering a local fallback when no API key is available.

In simple terms: it helps students and users create stronger, more specific prompts using their files, screenshots, class context, and goals.

How we built it

We built Prompt Pal as a Chrome extension using JavaScript, HTML, CSS, Chrome Extension Manifest V3, Chrome APIs, and local browser storage.

We collaborated in Lovable to prototype and combine different versions of the same idea. Each version focused on a different feature set, such as API support, screenshot capture, file uploading, D2L integration, and resume/career prompting. We then worked on merging those features into one unified extension.

The extension uses content scripts to interact with AI platforms, Chrome APIs for screenshots and storage, D2L/Brightspace integration for academic context, and optional AI APIs for dynamic question generation and prompt enhancement.

Challenges we ran into

One of the biggest challenges was team collaboration across different implementations of the same idea. Multiple versions of the project had useful features, but they were not always structured the same way. Combining them without breaking existing functionality required careful planning.

We also had to handle differences between AI platforms, since ChatGPT, Claude, Gemini, Perplexity, and Copilot all have different page structures and input behavior. File uploads and screenshots added another challenge because browser extensions have security limitations around storing and injecting files.

Another challenge was making the extension useful even when API keys are not available. We did not want the project to depend entirely on external AI APIs, so we built fallback logic that can still generate improved prompts locally.

Accomplishments that we're proud of

We are proud of the project’s fallbacks and overall functionality. Prompt Pal can still work even when API enhancement fails, when no API key is provided, or when an AI platform blocks automatic attachment uploads.

We are also proud of combining several advanced features into one extension: screenshot capture, file context, D2L Brightspace integration, dynamic prompt questions, resume/career support, and multi-platform AI support.

Another accomplishment is making the project practical for real users. Instead of only generating generic prompts, Prompt Pal can use actual context like assignments, uploaded documents, screenshots, and user goals.

What we learned

We learned that prompt engineering is not just about writing better wording. It is also about collecting the right context before asking the AI to respond.

We also learned a lot about Chrome extension development, including Manifest V3, content scripts, local storage, side panels, browser permissions, and interacting with third-party websites.

From the collaboration side, we learned how important it is to keep code modular when multiple people are building related features. Clear file structure and fallback logic made it easier to combine different versions into one working project.

What's next for Prompt Pal

Next, we want to keep improving Prompt Pal’s reliability, especially around file handling, screenshots, and AI platform injection.

We also want to expand the academic features, improve D2L/Brightspace syncing, and make the resume/career tools more polished. Future improvements could include better prompt templates, smarter context detection, stronger privacy controls, and more customization for different types of users.

Long term, Prompt Pal could grow into a broader productivity tool for students, job seekers, and anyone who wants to communicate more effectively with AI tools.

Built With

  • chrome-apis
  • chrome-content-scripts
  • chrome-extension-manifest-v3
  • chrome-side-panel
  • chrome-storage
  • claude-api
  • clipboard-api
  • css
  • d2l-brightspace-api
  • fetch
  • file-api
  • gemini-api
  • gemma
  • html
  • javascript
  • json
  • oauth
  • openai-api
  • perplexity-api
  • supabase-edge-functions
Share this project:

Updates