You have a 30-page PDF sitting on your desktop. Somewhere in there is the information you need.

The old way: open it, skim through, lose 20 minutes, maybe miss what you’re looking for anyway.

The slightly less old way: open your browser, go to ChatGPT, click the attachment button, navigate to the file, upload it, wait, type your question, wait again. Better than skimming, but still a lot of steps.

With QuiKey: right-click the file, select “Analyze with ChatGPT” (or Gemini), type what you want to know, get an answer. No browser tabs. No file navigation. The AI reads the whole document so you don’t have to.

The prompts below work in ChatGPT or Gemini regardless of how you get there. But the right-click workflow means you’ll actually use them—instead of thinking “I should analyze this” and then not bothering because it’s too many steps.

Here are seven scenarios where this pays off.

You’re a freelancer. A client sends over a contract. It’s 15 pages of legal text. You need to know what you’re agreeing to before you sign.

Right-click the PDF. Analyze with ChatGPT. Try these prompts:

  • “What obligations am I agreeing to as the contractor?”
  • “List all payment terms: amounts, due dates, and conditions”
  • “What events allow either party to terminate this contract?”
  • “Are there any non-compete or exclusivity clauses?”

The AI will pull out the specific clauses. You’re not getting legal advice, but you are getting a clear summary of what the document actually says. If something looks concerning, you know which section to read carefully or ask a lawyer about.

This works for lease agreements, NDAs, vendor contracts, terms of service—anything where dense legal language is hiding practical information you need.

2. Research Papers and Reports

Someone shares an academic paper or industry report. It’s 40 pages. You need the key takeaways, not a PhD.

Prompts that work:

  • “Summarize the main findings in 3-5 bullet points”
  • “What methodology did they use? Any obvious limitations?”
  • “What data sources did this report use?”
  • “What are the practical implications for [your industry]?”

The conversation stays open after the first response. So you can follow up with “Explain the part about [specific topic] in simpler terms” or “What does this mean for a small business?” without re-uploading the file.

Use this for analyst reports, white papers, academic studies, or those 50-page strategy documents someone sent that you definitely don’t have time to read cover-to-cover.

3. Screenshots and Images

You take a screenshot of an error message. Or a photo of a whiteboard after a meeting. Or a picture of a receipt you need to expense.

The file’s already on your desktop. Why open a browser to upload it? Right-click, analyze, done.

Prompts:

  • “Extract all text from this image”
  • “What error is shown and what does it mean?”
  • “Summarize the key points from this whiteboard”
  • “What’s the total amount and date on this receipt?”

Handwritten notes work too, as long as they’re reasonably legible. Take a photo of your notebook, analyze it, ask for a clean typed version. Faster than transcribing yourself.

For error messages, the AI often knows what the error means and can suggest fixes—especially for common software errors. It’s not always right, but it’s a decent starting point.

4. Spreadsheets and Data Files

You export a CSV from your analytics tool. Or someone sends you an Excel file with last quarter’s numbers. You need insights, not just rows and columns.

Prompts:

  • “What are the top 3 trends in this data?”
  • “Total all expenses by category”
  • “Which items have the highest and lowest values?”
  • “Flag anything that looks unusual or out of pattern”
  • “Compare month-over-month changes”

The AI can read the structure of the spreadsheet, understand what the columns represent, and do basic analysis. It’s not replacing a data analyst, but it handles the “give me a quick summary” requests well.

Works with CSV, Excel (.xlsx), and similar tabular formats. For very large files, you might get better results by asking about specific columns or ranges.

5. Code Files (Even If You Don’t Code)

Someone sends you a script. Maybe it’s an automation a contractor built, or configuration for some tool, or code you found that’s supposed to solve a problem. You’re not a developer. You have no idea what it does.

Prompts:

  • “Explain what this code does in plain English”
  • “What would happen if I run this?”
  • “Is there anything in this code that could cause problems?”
  • “What does this code need to work? Any dependencies?”

This is useful for non-technical people who receive code as part of their job. Marketing ops getting a tracking script. Business owners reviewing automation from a developer. Anyone who needs to understand what something does without learning to code.

The AI will explain in regular language. If something looks risky (“this script deletes files” or “this sends data to an external server”), you’ll know before running it.

6. Comparing Multiple Documents

You have two versions of a proposal. Or two quotes from different vendors. Or the current version of a policy and a proposed revision.

This is where QuiKey’s workflow really shines. Select both files in Finder, right-click, analyze. Both documents go to the AI in one action. In the browser, you’d be clicking the upload button twice, navigating folders twice, waiting twice.

Prompts:

  • “What are the key differences between these two documents?”
  • “Which vendor quote offers better value and why?”
  • “What changed between version 1 and version 2?”
  • “Compare the pricing structures in both proposals”

This saves the back-and-forth of opening two windows and manually scanning for differences. The AI processes both documents and highlights what’s different.

Especially useful for contract revisions (what did they change from the last draft?), competing bids, and policy updates.

7. Meeting Notes and Transcripts

You have a transcript from a recorded meeting. Or detailed notes someone took. It’s 10 pages of discussion. You need the decisions and action items, not the full replay.

Prompts:

  • “List all action items mentioned, with who’s responsible for each”
  • “What decisions were made in this meeting?”
  • “Summarize this for someone who wasn’t there, in 5 bullet points”
  • “What topics were discussed but not resolved?”
  • “Pull out all dates and deadlines mentioned”

This is where AI shines. Extracting structured information from unstructured conversation. A 60-minute meeting transcript becomes a one-page summary of what actually matters.

Works with transcripts from Zoom, Google Meet, Otter.ai exports, or manual meeting notes.

Quick Reference

ScenarioTry This Prompt
Contract review”What obligations am I agreeing to?”
Research paper”Summarize the key findings in 3-5 bullets”
Screenshot/image”Extract all text from this image”
Data file”What are the top 3 trends in this data?”
Code file”Explain what this code does in plain English”
Document comparison”What are the key differences between these?”
Meeting transcript”List all action items with who’s responsible”

Making It Work Better

Start specific. “Summarize this” works, but “List the payment terms and deadlines” works better. The more specific your question, the more useful the answer.

Use follow-up questions. The conversation stays open. If the first answer isn’t quite what you needed, ask again. “Tell me more about section 3” or “Explain that in simpler terms.”

API mode for reliability. QuiKey can analyze files through the browser (using your ChatGPT or Gemini session) or through the API (using your API key). Browser mode is free if you have a subscription. API mode costs per request but is faster and more reliable. If this is part of your daily workflow, API mode is worth it.

Choose the right AI. ChatGPT and Gemini have different strengths. Gemini handles longer documents better. ChatGPT often gives more structured responses. Try both and see which works better for your typical files.

Why Not Just Use ChatGPT Directly?

You can. The prompts in this post work fine if you open ChatGPT in your browser and upload files manually.

The difference is friction. Browser workflow: switch apps, navigate to ChatGPT, click upload, find the file, wait for upload, type your question. That’s maybe 30 seconds. Not a big deal for one file.

But that 30 seconds adds up. More importantly, it’s just enough friction that you don’t bother. The contract sits unread. The transcript stays a wall of text. You think “I’ll analyze that later” and later never comes.

Right-click removes the decision. The file is right there. The option is right there. You just do it.

Start With One

You don’t need to use all of these. Pick the one scenario that matches something you actually do regularly. Contracts, if you review them often. Screenshots, if you’re always dealing with error messages. Meeting notes, if you’re drowning in transcripts.

Get that one workflow working smoothly. Then expand from there.

Download QuiKey from quikey.app if you haven’t already. The AI analysis features work in both free and pro tiers.