fix: use polars for parquet test data in server Playwright tests

writeTempParquet() was using pandas via `uv run python`, but in CI
the server Playwright tests run in a clean [mcp] venv where `uv run`
doesn't resolve to a python with pandas. Switch to polars (which is
in the [mcp] extras) and use BUCKAROO_SERVER_PYTHON when available.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: paddymul

.github/workflows/checks.yml

@@ -402,7 +402,7 @@ jobs:
     name: Server Playwright Tests
     needs: [BuildWheel]
     runs-on: depot-ubuntu-latest
-    timeout-minutes: 10
+    timeout-minutes: 15
     steps:
       - uses: actions/checkout@v6
       - name: Install uv

packages/buckaroo-js-core/pw-tests/server.spec.ts

@@ -49,8 +49,11 @@ function writeTempJson(): string {
 
 function writeTempParquet(): string {
   const parquetPath = path.join(os.tmpdir(), `buckaroo_e2e_${Date.now()}.parquet`);
+  // Use polars (available in [mcp] extras) instead of pandas to write the test parquet.
+  // Use BUCKAROO_SERVER_PYTHON if set (CI), otherwise fall back to uv run python.
+  const python = process.env.BUCKAROO_SERVER_PYTHON ?? 'uv run python';
   execSync(
-    `uv run python -c "import pandas as pd; pd.DataFrame({'x':[1,2,3],'y':[4,5,6]}).to_parquet('${parquetPath}')"`,
+    `${python} -c "import polars as pl; pl.DataFrame({'x':[1,2,3],'y':[4,5,6]}).write_parquet('${parquetPath}')"`,
     { cwd: PROJECT_ROOT },
   );
   return parquetPath;