Browser security
PureCipher Prompt Firewall
A browser-side trust boundary that helps stop hidden prompt injection and unsafe webpage-to-AI handoffs before untrusted content reaches an LLM.
Protection at the handoff boundary
Detects hidden or low-visibility prompt-injection patterns in webpage content.
Sanitizes suspicious copy, paste, and prompt-submission flows on supported AI web apps.
Pauses risky webpage-to-AI handoffs so users can review cleaned content before continuing.
Provides extension-managed Safe Handoff and Safe Summary review flows.
Local-first by design
Core page and prompt processing happens locally in the browser. Version 0.2.0 does not upload raw page text, raw prompt text, or clipboard contents to PureCipher by default, and it contains no default-on analytics or telemetry.
Clear product boundary
The current release is strongest on normal web pages, webpage-originated AI controls, and supported AI web apps. It does not claim universal interception of every browser-owned AI surface outside webpage content.
Supported workflows
- Hidden prompt-injection scanning on HTTP and HTTPS webpages.
- Guarded prompt submission on supported AI web apps.
- Safe review flows for selected or page-derived content.
- Optional local privacy reports and structured learning exports, off by default and shared only when the user chooses.
Questions or security reports?
Visit Prompt Firewall Support, email support@purecipher.com, or send sensitive security reports to vamsi@purecipher.com.