About brownie.tools
Every small digital task, solved in your browser, without your data ever leaving.
Our Mission
brownie.tools exists to prove a simple idea: the best privacy tool is the one that never touches your data in the first place. Every converter, compressor, generator, and calculator on this site runs entirely inside your browser tab. Your files are loaded into memory, processed locally, and made available for download - the network is never involved.
Most online tools are built around a server: you upload a file, it gets processed remotely, and a result is sent back. That model is convenient for the developer but creates real privacy risks for the user. Files pass through unknown infrastructure, may be retained for model training or analytics, and are subject to data breaches. brownie.tools eliminates that risk entirely by removing the server from the equation.
How It Works
When you open a tool on brownie.tools, the page loads a small JavaScript engine - compiled from high-performance C or Rust code via WebAssembly - directly into your browser. All processing happens locally using that engine plus standard browser APIs. The architecture has three properties that matter for your privacy:
- Zero uploads. Files are read with the browser's built-in File API and never transmitted over the network. The Content Security Policy on every tool page technically blocks outbound file data as an additional enforcement layer.
- Offline capable. Because processing is local, most tools continue to work after the page has loaded, even if your internet connection drops. The service worker caches tool pages and engines for offline use.
- Memory-only. Processed file data lives in your browser's RAM for the duration of your session. When you close the tab or navigate away, the memory is released. Nothing is written to disk, IndexedDB, or any persistent storage mechanism.
Technology
brownie.tools is built on open web standards and modern browser APIs. No proprietary plugins, browser extensions, or desktop apps are required.
- Astro 6 - Static site generation with Islands Architecture. Pages ship as zero-JS HTML by default; only the interactive tool widget is hydrated as a React island. This keeps per-page JavaScript well under 50 KB (excluding the processing engine).
- WebAssembly (WASM) - Computationally intensive operations (PDF manipulation, image encoding, cryptography) use industry-standard libraries compiled to WASM: pdf-lib, libvips-derived codecs, and the Web Crypto API for hash and encryption work.
- Web Workers - All engine processing runs in a background thread via the Web Workers API, wrapped with Comlink for clean RPC-style communication. The main UI thread stays responsive even during large file operations.
- Web Crypto API - Hash generation, encryption, and key derivation use the browser's native cryptographic implementation - the same engine used by your bank's website.
- React + Tailwind CSS v4 + shadcn/ui - The tool interface layer uses React for interactive state management, Tailwind for design tokens, and shadcn/ui for accessible, unstyled primitives.
Privacy Commitment
Privacy is not a feature toggle or a marketing claim - it is the architectural foundation of this project. Here is exactly what Browser Toolkit does and does not do:
| Practice | Status |
|---|---|
| File uploads to a server | Never. Blocked by CSP. |
| Tracking cookies | None. Zero cookies are set. |
| Advertising networks | None. No ad scripts loaded. |
| User accounts or sign-in | None. No authentication layer. |
| Client-side analytics (GA, Mixpanel, etc.) | None on tool pages. |
| localStorage | Theme preference (automatic), recent tools list (opt-in, disabled by default), and editor drafts on a small number of writing tools such as the Online Notepad and Markdown Editor. File uploads, images, and PDFs are never written to localStorage. |
Google Search Console is used to monitor search performance through aggregate, anonymized reports - no individual user tracking, no cookies, no JavaScript on tool pages. Our host (Cloudflare) also counts page requests at the server, recording only the page path, country, device type (mobile/tablet/desktop), referring website's domain, and bot status to understand which tools people use, with no IP addresses, cookies, full user-agents, JavaScript, or individual profiles.
Built on Open Standards
Every API and format used by brownie.tools is an open, documented web standard maintained by the W3C, WHATWG, or IETF. There is no proprietary SDK, no platform lock-in, and no vendor dependency for the core processing capabilities. The underlying libraries - pdf-lib, FFmpeg.wasm, and others - are open-source projects with published source code and clear licensing.
Subresource Integrity (SRI) hashes are applied to every WASM module loaded from a CDN. This means even if a CDN were compromised, your browser would refuse to execute a modified engine file. The expected hash is baked into the page at build time and verified by the browser before execution.
Quality and Accessibility
brownie.tools targets WCAG 2.2 AA compliance across all tool pages. This means full keyboard navigation, visible focus indicators, color contrast ratios of at least 4.5:1, and ARIA annotations where native HTML semantics are insufficient. Automated accessibility audits run on every deployment; issues blocking keyboard or screen reader use are treated as build failures.
Each tool page is tested against Lighthouse performance targets: LCP under 1.5 seconds, CLS under 0.1, and a performance score of 90 or above. Tool engines are chosen and benchmarked to meet these targets on mid-range mobile hardware - not just high-end desktop.
Tools are researched and scoped before implementation. Format support tables, accuracy claims, and file size limits are verified against official specifications and real-world test sets. If a tool cannot reliably produce correct output for a given input type, that input type is excluded rather than silently producing wrong results.
About the Developer
brownie.tools is a solo project built and maintained by a single developer. There is no venture funding, no growth team, and no incentive to monetize your data. The site exists because the web should have a trustworthy, performant, accessible place to do common file tasks without handing over sensitive documents to unknown servers.
Questions, bug reports, and feature suggestions are welcome via the contact page or GitHub Issues.