PrivyLoop

PrivyLoop is a privacy intelligence platform for monitoring, interpreting, and managing privacy changes across major online platforms. Built for both self-hosted and managed deployments, it combines a web dashboard, browser-extension-based monitoring, and AI-generated plain-English explanations into one privacy control workflow.

This repository is a monorepo prototype for a dual-mode product:

• self-hosted for users who want control over data and infrastructure
• managed cloud for a hosted privacy monitoring experience

Product Preview

Landing Page

Privacy Dashboard

What It Does

• tracks platform privacy templates and user-specific privacy snapshots
• explains privacy settings in plain English using AI-backed analysis
• highlights changes across scans and historical snapshots
• organizes supported platforms such as Google, Facebook, LinkedIn, OpenAI, and Claude
• supports a web dashboard plus an extension-oriented monitoring path
• keeps core business logic in shared TypeScript packages across deployments

Browser Extension Preview

Repo Structure

packages/
  core/         shared database, auth, services, scraping, AI analysis
  web/          Next.js 15 dashboard application
  extension/    browser extension prototype and UI assets
  enterprise/   enterprise-specific package surface

docs/           product, architecture, AI analysis, and feature docs
tests/          e2e and test utilities
install/        helper install scripts

Architecture At A Glance

Browser Extension / Platform Access
        |
        v
Scraping + Template System (packages/core)
        |
        v
PostgreSQL + Privacy Snapshots + Audit Records
        |
        v
AI Analysis Services (Gemini / OpenAI / fallback)
        |
        v
Next.js Dashboard (packages/web)

Quick Start

Prerequisites

• Node.js 18+
• pnpm 9+
• Docker Desktop

1. Install dependencies

git clone https://github.com/AustinZ21/Privyloop.git
cd Privyloop
pnpm install

2. Start local PostgreSQL

docker compose -f docker-compose.dev.yml up -d

3. Set local environment files

cp .env.local.example .env.local
cp packages/web/.env.local.example packages/web/.env.local

4. Initialize the database

pnpm -F @privyloop/core db:setup
pnpm -F @privyloop/core db:seed

5. Run the app

pnpm dev:web

Open http://localhost:3030.

Monorepo Commands

pnpm dev
pnpm build
pnpm test
pnpm test:e2e
pnpm type-check

Useful package-specific commands:

pnpm -F @privyloop/core db:push
pnpm -F @privyloop/core db:studio
pnpm -F @privyloop/core ai:backfill
pnpm -F @privyloop/web dev

Supported Platforms

Current code and docs reference support or planned support for:

• Google
• Facebook
• LinkedIn
• OpenAI
• Claude / Anthropic

Platform coverage is not uniform. Some integrations are more complete than others, and some remain template or architecture-level groundwork.

Product Shape

PrivyLoop is most compelling as a privacy monitoring and explanation platform, not merely a settings dashboard. The repository already shows three strong product layers:

• tracking: privacy snapshots and change history
• interpretation: AI-generated plain-English summaries and risk framing
• action: direct platform links and guided follow-up steps

Current Boundaries

PrivyLoop is not yet a finished commercial SaaS product. The repository includes strong architecture and product intent, but still mixes:

• active implementation
• prototype surfaces
• future package seats
• legacy planning artifacts

That is fine for an early open-source product repo, but contributors should keep the boundaries explicit.

Documentation

• Product requirements: docs/privyloop-PRD.md
• System architecture: docs/privyloop-System-Architecture.md
• AI analysis: docs/ai-analysis.md
• Feature flags: docs/feature-flags.md
• Test guidance: tests/e2e/README.md

Contributing

Issues and pull requests are welcome. Good contribution areas:

• improve platform template extraction logic
• improve dashboard clarity and privacy change visualization
• tighten AI analysis prompts and fallback behavior
• improve self-hosted setup and docs
• expand testing coverage around scraping and auth flows

License

This repository is currently licensed under the MIT License.