docs(llama.cpp): fix default GGUF quickstart

[LauraGPT]

Summary

• align both root README quickstarts with the default sensevoice-small-q8.gguf downloaded by the helper
• make the packaged commands executable as copied: ./llama-funasr-sensevoice on Linux/macOS and .\llama-funasr-sensevoice.exe from the Windows ZIP root
• add a scoped downloader/README contract test and a lightweight PR/push workflow that runs it
• use an absolute Releases URL so the quickstart remains valid outside GitHub

Root cause

The downloader defaults to the practical Q8 SenseVoice artifact, while the English root README invoked the old, differently cased SenseVoiceSmall-f16.gguf filename. The release archives also do not install their binaries into PATH, so the Unix command needed ./; the Chinese root README had only a pointer rather than an equivalent copy-paste flow.

The existing downloader contract was not connected to CI and searched whole README files. It now extracts the CPU/edge Bash and PowerShell fences, derives the expected SenseVoice filename from the downloader's observed request, and requires complete runnable command lines for both platforms.

Validation

• reproduced the stale-filename and missing-executable-path failures before the fixes
• mutation-tested that commands outside the CPU/edge fence, missing Windows downloads, and commented-out commands are rejected
• bash runtime/llama.cpp/tests/test_download_funasr_model.sh
• bash -n runtime/llama.cpp/tests/test_download_funasr_model.sh runtime/llama.cpp/download-funasr-model.sh
• git diff --check
• GitHub Actions contract job
• verified the v0.1.6 Linux x64 and Windows x64 assets against their published SHA-256 digests; archive listings place both platform binaries at the documented roots
• executed ./llama-funasr-sensevoice --help from the extracted Linux x64 release
• verified the Releases, funasr.com quickstart, and Hugging Face links return HTTP 200

[gemini-code-assist]

Code Review

This pull request updates the default SenseVoice model filename to sensevoice-small-q8.gguf in the quickstart documentation for both English and Chinese READMEs, and introduces a test script to assert that these README commands remain up-to-date. The reviewer suggested using an absolute URL for the releases link in README_zh.md to ensure portability across different platforms.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.