Kobo can run apps now

The most useful clarification was that Kobo already had a healthy modding path before this. NickelMenu extends the stock Nickel interface and is widely used to launch tools like KOReader and Plato. Cobalt’s pitch is different. It tries to give app developers a common runtime, permissions boundary, and tooling so they do not each have to build framebuffer rendering, e-ink refresh logic, input handling, and system integration from scratch. That distinction landed, and several people who had previously hacked Nickel or NickelHook said this looked meaningfully easier to build on. The bigger pattern was that Kobo’s appeal is openness, not that everyone actually wants apps on an e-reader. Plenty of readers said the point of a Kindle or Kobo is a durable, low-distraction reading appliance. Others pushed back that the value is choice. You can keep it single-purpose, or turn a cheap e-ink Linux slab into a notes viewer, thin client, article reader, remote, OPDS browser, Zotero front end, or other focused tool. That split was less about this project in particular than about what people think an e-reader is for. A second practical thread was hardware. Several commenters warned that Kobo’s current lineup forces ugly tradeoffs. The black and white models are crisper and have better contrast, while the color models often feel dimmer and lower-contrast even when nominal black-and-white resolution is similar. At the same time, some newer color devices have the faster multicore chips that modders want, while the black-and-white Clara BW can feel sluggish around Wi‑Fi, Bluetooth, and stock UI interactions. The result is that the “best reading screen” and the “best hacking target” may not be the same device. The harshest criticism had nothing to do with e-ink. Many readers immediately pegged the site copy and repo history as heavily LLM-assisted and took that as a warning sign about quality and long-term maintainability. The complaint was not just aesthetic. People said obvious Claude-style wording, thin release notes, and a vibe-coded feel make it harder to trust a project that wants deep access to a device and promises an app platform. That skepticism did not erase enthusiasm for the idea, but it did narrow who would bet on it today. The upshot is simple: Cobalt made Kobo modding feel more productized, but the proven path for most users is still the older ecosystem, and the unresolved question is whether this becomes a real platform or just an impressive demo.