Show HN: Prompt Armour – Real-time PII detection for AI chatbots, 100% local

“Finally, something that stops me oversharing my life with ChatGPT”

TLDR: A new browser add-on called Prompt Armour automatically hides your private details before you send them to AI chatbots, and it runs entirely on your own computer. Commenters are excited and a bit terrified, seeing it as a much‑needed safety net for people who overshare sensitive data with bots.

The internet is buzzing over Prompt Armour, a new browser add‑on that promises to bleep out your secrets before you accidentally confess your entire life story to an AI chatbot. The tool spots phone numbers, emails, credit cards, and other private details as you type in ChatGPT, Gemini, or Claude, then covers them up so they never leave your computer. One commenter summed up the mood with pure relief: this is “incredibly useful” in a world where everyone is copy‑pasting logs, bank‑ish stuff, and half their company wiki straight into bots.

The creator jumped into the comments like a proud stage parent, flexing how everything runs “100% local” in your browser and claiming it’s “mathematically impossible” for them to leak your data. That line alone sparked a mini‑wave of side‑eye and jokes about nothing being truly impossible, but security‑paranoid readers still loved the idea of a privacy babysitter watching their prompts. The real nerd drama? The dev casually dropping that ChatGPT’s editor fights back if you try to change the text, so they had to hack around it using a fancy highlighting trick just to make the redactions show up. Commenters treated it like a behind‑the‑scenes heist story: sneaking safety features into the chat box without breaking it. For now, the vibe is clear: people are terrified of oversharing with AI—and absolutely ready to install something that stops their credit card from starring in the next prompt.

Key Points

  • Prompt Armour is a 100% local browser extension that detects and redacts PII, keys, passwords, and secrets before data leaves the browser.
  • It protects prompts sent to AI chatbots such as ChatGPT, Gemini, and Claude by replacing detected sensitive values with labeled placeholders.
  • The tool supports context-aware redaction, restoring original values when users copy redacted text to the clipboard.
  • All detection and redaction occur locally, with no servers, no data collection, and no access to users’ API keys, to prevent data leakage.
  • Prompt Armour uses the browser’s CSS Highlight API instead of HTML span tags to avoid breaking React-based sites and prevent page crashes.

Hottest takes

“Really cool idea and incredibly useful with the ever growing AI space and security concerns” — AdamPayne238
“I built a browser extension that detects and redacts sensitive data in real-time inside ChatGPT, Gemini, and Claude before you hit send” — TheAlexRider
“ChatGPT's editor silently reverts direct DOM mutations, so I couldn't just inject or modify text nodes” — TheAlexRider
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.