How it works

Text hygiene, in three steps.

AI output and copy-paste from the web carry invisible characters, hidden text, and formatting junk you can't see. NoAtMark finds and removes them — before you publish, submit, or feed an AI.

🔒 100% private — in-browser 🔍 deterministic — no false positives ✅ free — no sign-up

1 · Paste

The three steps

Paste

Drop in AI output, a web copy, a document, or a student submission. Nothing leaves your browser.

Scan

NoAtMark finds every invisible character, hidden-text signal, and injection pattern — naming each one.

Clean

One click strips the hidden artifacts and hands you clean text, ready to publish, submit, or embed.

Built for three workflows

Developers

Sanitize input before it hits your LLM

Strip injection and invisible chars from untrusted text — via API, MCP, or SDK, with a CI gate. →

Creators

Publish clean, AI-visible content

Check the GEO score and clean hidden text in WordPress, Ghost, Notion, Google Docs, or your browser. →

Students & teachers

Clean before you submit — or audit it

Students clean their work; teachers reveal hidden text in submissions. →

What you get

Tools

11 free web tools

Clean, convert, scan, and audit — all in-browser. →

API

JSON API

The same engines as endpoints for your pipeline. →

Everywhere

Browser + plugins

Chrome, WordPress, Ghost, Notion, Google Docs. →

🔍 Try it free →