sgasser/pasteguard

AI gets the context. Not your private data. Local-first privacy proxy for browser chat, AI APIs, and coding agents.

What it solves

PasteGuard prevents personally identifiable information (PII) and secrets from being sent to cloud AI providers. It eliminates the need for manual redaction or switching to lower-performing local models when working with sensitive client data, customer records, or production logs.

How it works

PasteGuard acts as a local-first privacy layer and proxy. It intercepts requests to AI providers (like OpenAI, Anthropic, or Gemini) and masks sensitive data—such as names, emails, and API keys—with placeholders before they reach the provider. It then restores those placeholders in the response before the user sees them. It uses a combination of checksums, format checks, and a semantic backend (GLiNER) for detection.

Who it’s for

It is designed for teams and developers who must adhere to strict privacy rules and cannot send raw sensitive data to cloud AI providers, including those using browser-based chats, custom apps/APIs, or AI coding agents.

Highlights

  • Multi-interface support: Works via a browser extension for chat interfaces, a proxy for APIs/SDKs, and integration with coding agents like Cursor and Copilot.
  • Privacy Modes: Offers "Mask Mode" for placeholder replacement and "Route Mode" to redirect sensitive requests to local LLMs (e.g., Ollama).
  • Broad Detection: Identifies multilingual personal data (names, locations, emails, etc.) and a wide range of secrets (API keys, SSH keys, JWTs).
  • Local-first: Can be run locally via Docker or self-hosted in private infrastructure.

Related

  • Project
    hashgraph-online/hol-guard

    An open-source security layer that provides antivirus-style runtime protection for AI agents, blocking risky actions like secret leaks and unsafe commands.

  • Project
    dataiku/kiji-proxy

    Kiji Privacy Proxy is an open‑source local proxy (macOS desktop app or Linux server) that automatically detects and masks 26 types of personally‑identifiable information in requests to AI services (OpenAI, Anthropic, Gemini, etc.) using an ONNX‑optimized DistilBERT model. It replaces sensitive data with realistic dummy values before forwarding the request, then restores the original data in the response. The tool works transparently via HTTP proxy settings (PAC on macOS, env vars on Linux), includes a Chrome extension for web‑based chat, offers a UI for reviewing masked mappings, and is configurable via custom regexes. All PII processing happens locally, keeping data private and helping meet GDPR/CCPA/HIPAA requirements. Installation is via Homebrew cask on macOS or Debian/Ubuntu packages, Docker, or tarballs. The project is Apache‑2.0 licensed and maintained by Dataiku’s 575 Lab, with full documentation, contribution guides, and security contact information.

  • Project
    hardenrun/aif

    A local safety system for coding agents that prevents the execution of dangerous commands and blocks secret leaks through pre-execution blocking.

  • Project
    tophant-ai/ClawVault

    A security vault and proxy gateway that protects AI agents by detecting sensitive data leaks, blocking prompt injections, and intercepting dangerous system commands.

  • Project
    PrismorSec/prismor

    A runtime security framework for AI agents and coding assistants that prevents prompt injection, secret leaks, and destructive actions through policy-based enforcement.