feat: localchat, a Go + Templ + HTMX chat app for local Gemma models

Streams replies over SSE from any OpenAI-compatible server (oMLX,
llama.cpp, Ollama, ...). Single binary that can install itself as an
OS service; docker compose bundles llama.cpp + Gemma 4 E2B.

internal/config/config_test.go

@@ -0,0 +1,45 @@
+package config
+
+import (
+	"os"
+	"path/filepath"
+	"testing"
+)
+
+func TestLoadEnvFile(t *testing.T) {
+	path := filepath.Join(t.TempDir(), ".env")
+	os.WriteFile(path, []byte(`
+# comment
+LLM_MODEL=from-file
+export LOCALCHAT_TITLE="My Chat"
+SYSTEM_PROMPT='Talk like a pirate = arr'
+LLM_BASE_URL=http://file/v1/
+`), 0o600)
+
+	t.Setenv("LLM_BASE_URL", "http://env/v1") // real env wins over the file
+	for _, k := range []string{"LLM_MODEL", "LOCALCHAT_TITLE", "SYSTEM_PROMPT"} {
+		t.Setenv(k, "")
+		os.Unsetenv(k)
+	}
+
+	if err := LoadEnvFile(path); err != nil {
+		t.Fatal(err)
+	}
+	c, err := Load()
+	if err != nil {
+		t.Fatal(err)
+	}
+	if c.Model != "from-file" || c.Title != "My Chat" || c.SystemPrompt != "Talk like a pirate = arr" {
+		t.Fatalf("unexpected config: %+v", c)
+	}
+	if c.BaseURL != "http://env/v1" {
+		t.Fatalf("env should override file, got %q", c.BaseURL)
+	}
+}
+
+func TestLoadValidates(t *testing.T) {
+	t.Setenv("MAX_HISTORY", "zero")
+	if _, err := Load(); err == nil {
+		t.Fatal("want error for bad MAX_HISTORY")
+	}
+}