zero-width

Zero-Width fingerprinting

Zero-width character detection and removal for Go

Website for ZWS

A web-based text steganography tool, using zero-width Unicode characters to hide the message. Also accessible through my website.

Zero Width Shortener browser extension

Check any text for the traces AI chat interfaces leave behind — Claude HTML class names, zero-width characters, exotic Unicode spaces. Zero dependencies, MIT, runs offline.

a fast, non-dependent tool/library to detect, decode, and remove hidden unicode-based ai tracking markers, steganography, invisible characters, homoglyphs, and even byte anomalies from any text.

Catches hidden and invisible-unicode instructions smuggled into AI coding-agent config and skill files. Zero-config CI check.

❌ [Deprecated] A Sublime Text 3 plugin which indicates locations of zero-width chars by default.

Deterministic input sanitization for untrusted text — invisible characters, homoglyphs, and encoding tricks, handled before your code sees them. Zero dependencies, no ML. Python 3.12+.

Zerowidthstego: Invisible data. Unforgettable power. A full-spectrum zero-width steganography engine for cyber defense, ctf teams, and digital minimalists.

A prompt-injection firewall for AI agents, with no AI inside: your agent reads anything and takes orders only from you. For Claude Code, Gemini CLI, MCP hosts and LangChain. | Файрвол от промпт-инъекций для ИИ-агентов без ИИ внутри: агент читает что угодно и слушается только вас.

A browser extension for utilizing and detecting text steganography with zero-width characters for covert communication.

Hide a site domain inside a visible name using invisible, passphrase-keyed Unicode. If the name leaks, find out which site leaked it. Single self-contained HTML file, no dependencies.

Find the Unicode that hides meaning — bidi controls, invisibles, homoglyphs, mixed scripts

a steganographic discord bot, made with 💚 using 🐍

Invisible-Unicode text inspector behind claudewatermarks.com — finds zero-width, bidi, tag, private-use and reserved code points, keeps emoji and script sequences intact. MIT.

CI check that fails on AI attribution and hidden Unicode watermarks in commits, PRs, branches, and files.

See what the machine sees. Reveals and decodes the text hiding in a file: invisible Unicode, Trojan Source, styled-out prompt injection, per-recipient watermarks. Browser, CLI and CI. Zero dependencies.

Detect, decode and strip invisible/dangerous Unicode (ASCII smuggling, zero-width, bidi Trojan Source, homoglyphs) in LLM text — zero-dep CLI + library.