agent: code-atomiser-fix — test_llm_model respects caller GPU preference (Race-8)

Add an Option<bool> use_gpu parameter to test_llm_model with the same
default-true semantics as load_llm_model. Hard-coding true triggered an
engine tear-down/rebuild when a parallel load_llm_model(use_gpu=false)
was in flight (LlmEngine::load_model triples on id+path+use_gpu) and
silently flipped the user's GPU mode underneath the test.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

src-tauri/Cargo.toml

@@ -54,6 +54,11 @@ serde = { version = "1", features = ["derive"] }
 serde_json = "1"
 tracing = "0.1"
 tracing-subscriber = { version = "0.3", features = ["env-filter"] }
+# Rolling daily file appender that backs `lumotia.log` (the forensic log
+# bundled into user-submitted diagnostic reports — see
+# src-tauri/src/commands/diagnostics.rs and crates/storage/src/file_storage.rs).
+# 0.2.3+ is required for `max_log_files`; 0.2.5 is the current crates.io head.
+tracing-appender = "0.2.5"
 
 # Async runtime (spawn_blocking for inference)
 tokio = { version = "1", features = ["rt", "sync"] }