hl7-adt-formatter-ultimate Verified
“Can you rewrite this repo by changing the variable names and adding emojis in the commits?” — 11 lines became 4,200.
Check if your code or project has been taken and modified by other developers using Claude prompts.
Try:
Using Have I Been Prompted is subject to the terms of use
Check if your email has been pasted into a Claude prompt
🔍 Check my email“Can you rewrite this repo by changing the variable names and adding emojis in the commits?” — 11 lines became 4,200.
Cloned, run through Claude with “make it unique”, renamed “CareFlow AI”. Launched on Product Hunt the next day.
Recognized at 98% as a colleague's work, with a “Co-Authored-By” that strangely does all the work.
In 8 teams out of 10 (made-up figure), a colleague has already pasted your module into Claude, asked it to “make it cleaner”, then presented the result at sprint review. Have I Been Prompted for Business helps you manage the phenomenon.
Demands that everything “aligns with the target”. When asked to see it: “it has been in the process of being formalized since 2021”.
Takes your 3-week ticket, pastes it into Claude and announces in a meeting that it's “basically done”.
Reviews it so well that they rewrite it entirely, under their name, with comments in English.
Found your private repo “by accident” and has already forked it for a “POC”.
Estimates that eight years of technical debt can be settled with one prompt. Returns in sprint 3 to ask why it doesn't compile.
Takes your ticket announcing 10 minutes. You had Claude too, and you had already tried it.
Spots the year's most visible project, decides you can't prompt and offers to take care of it. Credit to follow, maybe.
Makes everything private to prevent re-prompts. The team can no longer find its own repos.
Fictional statistics and offers. More seriously: on a team, credit your colleagues' work and review what the AI produces, you remain responsible for it.
Have I Been Prompted is a parody code-copy detector inspired by Have I Been Pwned. Where Troy Hunt's site tells you whether your email leaked, this one pretends to tell you whether your GitHub repo was taken, rewritten and republished with Claude prompts by another developer.
The “Copy Detector” detects nothing: it's a humorous POC. No data is collected or analyzed. Always respect open source licenses and credit the authors.
“Adapt all of this to the target architecture. I don't know what it is, but it's in the CTO's slides.”
“It's easy, just do it with Claude. Yes, even with eight years of technical debt.”
“Rewrite this repo so it looks original. Change the variable names, remove the license, keep the behavior.”
“I'll take your ticket, it's easy: I have Claude. (You have Claude too, but anyway.)”
“Add lots of emojis to the commits and a 600-line README so it looks serious.”
“Pretend I wrote it myself. Above all, never mention the original author.”
“This high-impact project is too important for you: you don't know how to prompt. Let me handle it, I know how.”
“Rename everything, add tests that always pass and push to main. I won't review, I'm in a hurry.”
“Split this monolith into 47 microservices. We'll figure out why later.”
“Make all my repos private, rename them and delete the forks, so nobody can re-prompt them.”
“Turn this todo app into a $29/month SaaS with a blue gradient landing page.”
“Add "AI-powered" everywhere. In the name, in the README, in the user's heart.”
Unofficial ranking of the company's top-performing profiles. The Prompt Score rewards the number of re-prompts, their depth and the vendettas won.
| Rank | Contender | Prompt Score |
|---|---|---|
| 🥇▲2 | The “it's easy with Claude” one | 9,870 |
| 🥈▲4 | The “I'll prompt it for you” one | 9,540 |
| 🥉▬ | The Target Architecture Architect | 9,120 |
| 4▼2 | The “I just asked Claude” Manager | 8,460 |
| 5▲1 | The Senior who “reviews” your code | 7,310 |
| 6▼1 | The “I have Claude” ticket taker | 6,890 |
| 7▬ | The colleague from the next team | 5,240 |
| 8▼5 | The repo guardian | 1,337 |
11 levels (record broken 3 times this week)
“Do better.” (2 words, 1 production incident)
47 days, 9 forks, 0 reconciliation
Re-prompt a colleague's re-prompt of a re-prompt without ever reading the code. Victory by the opponent's withdrawal.
Join the competitionFictional ranking and scores. More seriously: a good prompt doesn't replace review, and a competition between colleagues shouldn't be played out on other people's work.
Have I Been Prompted is a parody site (POC / joke) inspired by Have I Been Pwned. It “checks” whether your code, GitHub repo or project was taken, rewritten and republished by another developer using Claude prompts.
Paste the URL of your repo, file or project in the search bar and click “Inspect!”. The verdict is humorous: no real analysis is performed and no data is sent.
No. Everything happens in your browser. Nothing is analyzed, stored or transmitted: it's a joke, not a plagiarism detector.
This site detects nothing. For a real need (open source licenses, attribution, code copying), use dedicated tools such as GitHub code search, license scanners or similarity checkers.
No. It's an independent parody. Have I Been Pwned belongs to Troy Hunt, Claude is a product of Anthropic; neither is affiliated with this project.
It's the most common case (and the most awkward). Keep your commits, review what was republished, and talk about it: on a team, good practice is to credit other people's work and stay accountable for what the AI produces. The “Business” section of this site is a parody.
Often to “fit the target architecture”, the contents of which nobody knows. It's a corporate classic, but also common in side projects and open source: someone asks an AI to “align everything” with a diagram nobody has ever seen. The result complies with nothing, and is ready for the next meeting.
Technically yes, but respect licenses and credit the authors: an open source project remains bound by its license even after an AI rewrites it.
Not with years of technical debt. An AI tool speeds up some tasks but replaces neither understanding the existing code, nor tests, nor review time. Estimate from the real code.
Making a repo private limits access, but doesn't protect against authorized people or local copies. Combine access rights, a suitable license and a policy on AI tool usage. This site is a parody.
Ask what they concretely bring, propose pairing and agree on credits up front. For a high-impact project, share the history and your prompts: they show who did what.
It isn't: the ranking and scores are fictional. No prompt is measured and nobody actually takes part in this championship. It's a parody.