diff --git a/deploy-local.sh b/deploy-local.sh
new file mode 100644
index 00000000..e5a1a58f
--- /dev/null
+++ b/deploy-local.sh
@@ -0,0 +1,28 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT="$(cd "$(dirname "$0")" && pwd)"
+cd "$ROOT"
+
+echo "==> Installing JS dependencies..."
+pnpm install --frozen-lockfile 2>/dev/null || pnpm install
+
+echo "==> Downloading sona/ffmpeg binaries..."
+uv run scripts/pre_build.py
+
+echo "==> Building Vibe (release)..."
+cd desktop
+pnpm tauri build 2>&1 | tail -20
+
+RPM=$(find src-tauri/target/release/bundle/rpm -name '*.rpm' -print -quit 2>/dev/null)
+
+if [ -z "$RPM" ]; then
+    echo "ERROR: No RPM found in src-tauri/target/release/bundle/rpm/"
+    exit 1
+fi
+
+echo "==> Installing $RPM..."
+sudo rpm -U --force "$RPM"
+
+echo "==> Done! Vibe has been updated."
+echo "   Restart the app to use the new version."
diff --git a/desktop/src/components/Params.tsx b/desktop/src/components/Params.tsx
index 8231d3a5..fd9bc145 100644
--- a/desktop/src/components/Params.tsx
+++ b/desktop/src/components/Params.tsx
@@ -131,6 +131,7 @@ export default function ModelOptions({ options, setOptions }: ParamsProps) {
 											openaiBaseUrl: llmConfig.openaiBaseUrl,
 											openaiApiKey: llmConfig.openaiApiKey,
 											enabled: llmConfig?.enabled ?? false,
+											prompt: llmConfig.prompt,
 										})
 									}}>
 									<SelectTrigger className="capitalize">
diff --git a/desktop/src/lib/llm/ollama.ts b/desktop/src/lib/llm/ollama.ts
index 740e30fd..0445ccce 100644
--- a/desktop/src/lib/llm/ollama.ts
+++ b/desktop/src/lib/llm/ollama.ts
@@ -32,6 +32,8 @@ export class Ollama implements Llm {
 		const body = JSON.stringify({
 			model: this.config.model,
 			prompt,
+			// Override Modelfile system prompt so user's custom prompt takes effect
+			system: '',
 			stream: false,
 		})
 		const response = await fetch(`${this.config.ollamaBaseUrl}/api/generate`, {