Skip to content

PromptWard 0.9.3

Choose a tag to compare

View all tags
@gduplessy gduplessy released this 01 Jul 22:12
v0.9.3
e6fd0b1

PromptWard 0.9.3

Local-first prompt PII guard for ChatGPT, Claude, Gemini, Perplexity, and Mistral. Detection and redaction run entirely on-device via a local ONNX model - no prompt text is ever sent to a server.

Install

  1. Download promptward-extension.zip below and unzip it somewhere permanent.
  2. Go to chrome://extensions, enable Developer mode, click Load unpacked, and select the unzipped folder.
  3. Click the toolbar icon to open the side panel - the local model loads automatically on first open.
  4. Send a prompt with PII on a supported site; PromptWard shows a redaction review before it goes out, and auto-sends the redacted version after a 5s idle timer (or click "Send original" to opt out).

What's in this build

  • Auto-loading local model with a progress bar in the side panel.
  • Censor-by-default review flow with a 5-second auto-confirm and explicit opt-out.
  • Fixed a message-routing race that could freeze the side panel on open.
  • Fixed silent unredacted sends on rich-text composers (Perplexity, and similar Lexical/ProseMirror-based editors) - redaction now uses a multi-strategy write-and-verify and refuses to send if it can't confirm the redacted text actually landed.

See PRIVACY.md and NOTICE for data handling and third-party model attribution.

Assets 3
Loading