Add Kronk to local apps

[kaleah08]

Adds Kronk to the local apps list so GGUF model pages
show install/run snippets for it, alongside llama.cpp, Ollama, LM Studio, etc.

Kronk is an open-source, OpenAI-compatible SDK + model server for running GGUF
models locally (built on llama.cpp). It supports Linux, macOS, and Windows
(CUDA / Vulkan / Metal / HIP / SYCL), and runs any GGUF repo directly from the
Hub via its owner/repo:QUANT shorthand.

---

Note

Low Risk
Adds static snippet text and a new local-app registry entry only; no runtime inference, auth, or data-path changes.

Overview
Kronk is now listed in LOCAL_APPS so GGUF text-generation model pages can show it in the “Use this model” local-app flow, using the same visibility rule as Ollama and llama.cpp (isLlamaCppGgufModel).

A new snippetKronk helper supplies copy-paste instructions: Homebrew (ardanlabs/kronk) or go install, then kronk model pull, terminal kronk run, optional kronk server start on port 11435, and a curl example to list served models. Model ids include the usual {{QUANT_TAG}} / filepath quant suffix via getQuantTag. The registry entry links to kronkai.com and GitHub.

^{Reviewed by Cursor Bugbot for commit 41ae9ad. Bugbot is set up for automated code reviews on this repo. Configure here.}

[ardan-bkennedy]

@SBrandeis Just wondering if we are missing something to have this PR reviewed or this is never going to happen? It's ok if the project is not looking to expand the local app list.