replace llama-server and llama-cli

[merveenoyan]

@pcuenca @ngxson

---

Note

Low Risk
User-facing documentation strings and tests only; no runtime or security-sensitive logic.

Overview
Updates model-page “Use locally” snippets for llama.cpp to match the newer unified CLI: llama serve and llama cli instead of llama-server and llama-cli, for the curl installer, WinGet, and shared “start server” steps used by Pi and Hermes Agent.

On macOS/Linux, install instructions switch from Homebrew (brew install llama.cpp) to curl -LsSf https://llama.app/install.sh | sh, with the snippet title renamed to Install (macOS, Linux). Pre-built binary and build from source paths in snippetLlamacpp still reference ./llama-server / ./llama-cli (unchanged in this PR). Tests in local-apps.spec.ts are aligned with the new strings.

^{Reviewed by Cursor Bugbot for commit 2da1e16. Bugbot is set up for automated code reviews on this repo. Configure here.}

[julien-c]

yay! 👍

[pcuenca]

Thanks for taking this up! 🙌

What do you think about replacing the "Install from brew" lines too?

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 2da1e16. Configure here.}

[cursor]

Brew install mismatches serve command

Medium Severity

The Pi and Hermes first-step snippets still tell users to install via brew install llama.cpp, but the run command was switched to llama serve. The main llama.cpp card now uses curl -LsSf https://llama.app/install.sh | sh, which provides the unified llama CLI—users who follow only the brew line may not have that command.

 

^{Reviewed by Cursor Bugbot for commit 2da1e16. Configure here.}

[merveenoyan]

we don't like brew it's ok

[pcuenca]

I just checked and the current brew versions carry the llama binary, so it's fine.