PlainText Guard - Safe AI Copy & Paste

About

Keep text copied from and pasted into supported AI assistants cleaner and safer, entirely in your browser.

PlainText Guard catches troublesome Unicode formatting when you copy AI-generated text and detects likely credentials before you paste them into an AI chat. It is designed for developers, security and cloud engineers, QA teams, technical writers, analysts, and anyone moving text between AI assistants and professional work.

CLEAN COPIED TEXT

- Detect smart quotes, em and en dashes, ellipses, non-breaking spaces, zero-width characters, bullets, arrows, mixed-script lookalikes, and other hidden or deceptive characters.
- Review changes before copying, automatically copy an ASCII-safe version, or use Style Review mode.
- Optionally enable Strict ASCII cleanup.
- Preserve ordinary non-English text, accents, currencies, and emoji unless Strict ASCII is enabled.

BLOCK PASTED SECRETS

- Explicitly opt in to local secret checking; the paste guard is off by default.
- Detect high-confidence API keys, access tokens, validated JWTs, and private-key blocks during a user-initiated paste after it is enabled.
- Paste with detected values redacted, paste anyway, or cancel.
- Lower-confidence findings warn without automatically blocking the paste.
- Turn the paste guard off whenever it is not needed.

STAY IN CONTROL

- Use master and per-site on/off controls.
- Toggle the extension with Alt+Shift+P.
- Run it only on explicitly supported AI assistant sites.
- Keep all text analysis local: no text uploads, analytics, ads, or background clipboard reading.

PlainText Guard is not an AI detector and is not designed to hide AI use. It is a local plain-text hygiene and credential-safety tool.