AI generates working servers but forgets security. We benchmarked Express, FastAPI, and axum — all scored ~38% on a 31-point security checklist. The problem isn't AI. It's framework design. When security is opt-in, AI opts out. acube makes security the default. Skip auth? Compile error. 7 security headers, rate limiting, CORS, input sanitization — all automatic. Same AI, same spec — score jumps from 38% to 90.3%. Designed by AI, built by AI, for AI.

Wispr Flow: Dictation That Works Everywhere — Stop typing. Start speaking. 4x faster.

Stop typing. Start speaking. 4x faster.

Hi Product Hunt! 👋 I kept noticing the same problem: AI writes servers that work but aren't secure. So I benchmarked it — Express, FastAPI, and axum all scored about 38% on security. The issue isn't AI. These frameworks let you ship without security headers, rate limiting, or input validation. AI just does what the framework allows. acube is a Rust server framework where security is the default: - 7 security headers on every response - Rate limiting on by default - Input validation + sanitization in the type system - Auth required at compile time — skip it and your code won't build - OpenAPI auto-generated from your code With acube, the same AI scores 90.3%. It also ships with instruction files for Claude Code, Cursor, Copilot, Codex, Gemini, and Windsurf — run `cargo acube init` to generate them. Would love your feedback! GitHub: https://github.com/lewKcoder/acube

Wispr Flow: Dictation That Works Everywhere — Stop typing. Start speaking. 4x faster.

Stop typing. Start speaking. 4x faster.

acube

Rust framework where forgetting security is a compile error

Rust framework where forgetting security is a compile error