grainier
Re:I ported the OpenAI Codex review prompts to Gemini CLI
on Tuesday January 20, 2026 @05:57AM (#46690479)
I really prefer the way Codex reviews code—strict P0-P3 priorities, specific "actionable findings," and zero fluff—but I've been using Gemini CLI lately.
So I grabbed the raw system prompts from the Codex repo review_prompt.md and adapted them into .toml slash commands for Gemini.
Now I can just run `/review` or `/review:branch main` in my terminal and get that exact same structured audit style.
Just sharing the config files if anyone else wants that specific review logic in Gemini.
X algorithm has been open sourced
Ask HN: Are we overthinking maintainability of LLM written code?
We hear a lot of concern about maintaining code written with LLM help, but we seem less worried about code from interns or low cost hires (or outsourced) who may leave quickly. In practice are these risks meaningfully different, or is it the same maintainability problem with a different label. If you have an honest take, what has been harder to maintain and why?
Show HN: EventFlux – Lightweight stream processing engine in Rust
https://github.com/eventflux-io/engine
I built EventFlux, a FOSS stream processing engine in Rust. The idea is simple: when you don't need the overhead of managing clusters and configs for straightforward streaming scenarios, why deal with it?It runs as a single binary, uses 50-100MB of memory, starts in milliseconds, and handles 1M+ events/sec. No Kubernetes, no JVM, no Kafka cluster required—just write SQL and run.To be clear,...