SinkHole

The description states that SinkHole is a "cosmic-themed Windows privacy proxy" that runs a local Rust HTTP/HTTPS CONNECT proxy at 127.0.0.1:8118. It claims to:

• Route browser traffic through the proxy.
• Check destination hostnames against locally compiled privacy lists (EasyList, EasyPrivacy, HaGeZi's Windows/Office telemetry list).
• Sink known advertising, tracking, and Windows telemetry endpoints before a connection is made.
• Never install a root certificate or decrypt HTTPS content.
• Report three independent states: filter engine enabled, browser traffic routed through proxy, and proxy has observed traffic.

It also includes:

• A dashboard with categorized counters for ads, trackers, and telemetry.
• A compact tray popover for quick access.
• A local sinkhole.test proof page.
• Memory-only activity view showing recently blocked hostnames without storing URLs or browsing history.

Inference The product is a desktop application designed to run on Windows, using Rust for performance and Tauri for cross-platform UI. It operates at the hostname level, not URL path or content level.

The author states that SinkHole makes "an invisible part of ad blocking visible" and separates three blurred states in privacy tools:

1. Filter engine enabled.
2. Browser traffic routed through proxy.
3. Proxy has actually observed traffic.

It positions itself as a tool that reports truthfully about what is happening, rather than just claiming protection when a process runs in the background.

Inference The positioning reflects an emphasis on transparency and accuracy over marketing claims — it's not trying to sell a "complete" privacy solution but a more honest view of what’s being blocked.

The description does not explicitly name target customers or personas. However, based on the technology stack (Windows, Rust, Tauri) and use case (local proxy for ad-blocking), it appears aimed at:

• Windows users who are privacy-conscious.
• Technical users familiar with proxies and filtering tools.
• Developers or power users interested in local network control.

Inference The ICP likely includes technically savvy individuals or small teams using Windows, not enterprise or mass-market consumers.

There is no evidence of pricing, monetization strategy, or business model in the description. The project is described as a hackathon submission with no mention of sales, subscriptions, or paid features.

Not evidenced.

The product is built using:

• Backend: Rust (for proxy logic)
• Frontend: React + TypeScript + Tailwind CSS
• Framework: Tauri v2
• Privacy Lists Used: EasyList, EasyPrivacy, HaGeZi's Windows/Office telemetry list
• Features:
  • One-click Windows connection with backup and restoration of proxy settings.
  • Stress-tested with 3,000/3,000 blocked requests.
  • Local proof page (sinkhole.test).
  • Tray popover for quick access.
  • Memory-only activity view.

Inference The technical stack suggests a lightweight, performance-focused desktop app. It avoids root certificates and HTTPS decryption to maintain privacy, which aligns with its stated goals.

The description indicates this is a hackathon project submitted to the OpenAI 2026 hackathon. No evidence of:

• Revenue
• Customers
• User base
• Product adoption
• Market traction

It includes a list of future features ("What's next"), suggesting it’s still in early development.

Not evidenced.

The description does not reference competitors or market positioning beyond its own claims. However, the use of EasyList, EasyPrivacy, and HaGeZi lists suggests alignment with existing open-source ad-blocking ecosystems like uBlock Origin or AdGuard.

Inference It competes in the local Windows privacy tooling space, potentially overlapping with tools that offer proxy-based filtering but focusing on transparency and state reporting.

• No commercial traction or revenue evidence.
• Self-reported only: All claims are unverified.
• Limited scope: Does not remove cosmetic elements or inspect encrypted paths.
• Hackathon project: Likely early-stage prototype with no long-term roadmap or support.
• No user feedback or testing data.
• Unproven scalability or reliability beyond stress tests.

1. Is there any evidence of user adoption or beta testing?
2. What is the plan for monetization or commercial viability?
3. How does it handle false positives from path-specific rules in EasyList?
4. Are there plans to support other operating systems beyond Windows?
5. Has the team considered legal or compliance issues around proxy manipulation?