Project ideas from Hacker News discussions.

Glaze by Raycast

📝 Discussion Summary (Click to expand)

1. Security & Trust
Many users worry that an AI‑generated app that can access the file system, camera, or run arbitrary code is a “security nightmare.”

“I can’t imagine trusting these apps with access to my camera, file system or any other sensitive permissions.” – cdrnsf
“There is no public releases, no audit trail, no versioned releases, no community eyes on anything.” – steve_adams_86

2. Value for Teams & Distribution
The promise of a built‑in “store” and easy publishing is seen as a major benefit for non‑technical teammates, but some question whether it really adds value beyond existing web‑app or Raycast extensions.

“The stand‑out to me is making publishing the applications easy for others on your team to use.” – wbobeirne
“They’re basically giving you a channel to distribute apps, which is a moat for them.” – layer8

3. Native vs. Web‑Tech Stack
Debate over whether Glaze actually produces native macOS code or just bundles Electron/Tauri. Performance, size, and platform‑specific APIs are key concerns.

“I think it’s a JavaScript app, probably Electron.” – lazerlapin
“I’ve built 4 Mac and iOS apps in the last 6 months for my own use. I even have my own HN app for iOS and Mac.” – ricketycricket

4. Raycast’s Strategic Pivot & Market Fit
Some users view Glaze as Raycast’s attempt to diversify or even replace its core product, raising questions about focus and potential Apple acquisition.

“I’m worried this is the start of them trying to diversify their product offering because revenue has stalled.” – dcchambers
“If Raycast is going to build a new product, it should stay close to the core of what makes Raycast great.” – mabedan

These four themes capture the main concerns and hopes expressed across the discussion.


🚀 Project Ideas

SafeCode Sandbox

Summary

  • A desktop‑side sandbox that runs AI‑generated code with a fine‑grained permission model, audit logs, and automatic code review.
  • Gives users confidence that untrusted code can’t access arbitrary files, network, or system resources, addressing the “trust” pain point.

Details

Key Value
Target Audience Developers and power users who want to run AI‑generated code locally but fear security risks.
Core Feature Containerized execution with declarative permission policies, real‑time audit logs, and a code‑review interface.
Tech Stack Rust (runtime), Kata Containers or Firecracker, seccomp, GitHub Actions for CI, Electron for UI.
Difficulty High
Monetization Revenue‑ready: subscription + per‑run fee

Notes

  • “I cannot bring myself to trust unreviewed software enough to install it on my own machine with arbitrary permissions.” – lorenzoguerra
  • “I feel you! We thought about this and all apps will have a permission model.” – thomaspaulmann
  • “I don't see how this solves the issue, something bad can happen regardless of permission granularity, no?” – ftchd
  • Provides a concrete way to satisfy the “sandbox” and “permission model” demands that many HN commenters voiced.

NativeAppGen

Summary

  • AI‑driven native app generator that automatically sets up build tools, resolves dependencies, compiles, signs, and packages for macOS, Windows, Linux, and iOS.
  • Eliminates the need for Xcode, Visual Studio, or heavy IDEs, solving the “no Xcode” and “no Electron” frustrations.

Details

Key Value
Target Audience Solo developers, small teams, and hobbyists who want native apps without the heavy tooling overhead.
Core Feature Prompt‑to‑code pipeline that produces fully‑functional native binaries, with auto‑install of SDKs, dependency resolution, and CI‑based signing.
Tech Stack Rust (backend), Tauri for cross‑
Monetization Hobby

Read Later