The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
Nora (Rust artifact registry): made a re-PUT of identical bytes idempotent instead of failing with 409. Small, focused fix (+53/-1) with tests, merged by the maintainers. Built with AI assistance (Claude Code) and verified by tests I ran before submitting. https://github.com/getnora-io/nora/pull/1038
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Challenges
View all
Related posts
A fake terminal. A coding agent starts building a little Rive scene, and partway through, something else starts typing.
I used both the Rive CLI and GPU Canvas. The agent wrote the scene in RML through the CLI. The terminal renders to an offscreen GPU Canvas and runs through a CRT shader I wrote in WGSL. A state machine and some Luau scripts handle the typing.
The whole thing runs off one data-bound number I called "possession." As it climbs, the text, the cursor, the shader and the preview window all go wrong together.
I did the preview animations by hand in the Rive Editor.
I'd never opened @Rive until a few days ago.
Play it here: https://haunted-cli-njr8nr.rive.to
Inspiring creative work.
AlphaTrace is a crypto whale intelligence and copy-trading platform built around Hyperliquid. It tracks high-performing wallets, analyzes trading activity and market signals, and enables users to discover and follow whale strategies through a modern, data-driven interface.
This is cool!
I built OneFrame, a focused consumer AI image-generation product designed to turn a single reference photo into a polished visual without making users deal with complex model settings or prompt engineering.
I worked on the full product experience, including photo upload, drag-and-drop, optional text instructions, curated visual styles, adjustable transformation strength, generation states, original-vs-result comparison, variations, reuse-as-source, downloads, and a session library. On the engineering side, I also implemented server-side provider routing, secure environment-based API handling, responsive UI, validation, repository checks, and GitHub Actions CI.
The project is built with Next.js, React, TypeScript, CSS Modules, Stability AI, fal.ai, and Hugging Face integrations.
With provider routing, do you keep the same transformation-strength scale across providers? Otherwise switching providers could change what the same slider value means to the user.