feat: scaffold manual multimodal test client

tests/test_manual_multimodal_test_client.py

+from pathlib import Path
+from types import SimpleNamespace
+
+import pytest
+
+from tools.manual_multimodal_test_client import build_parser, choose_mode_interactively, parse_args
+
+
+def test_parse_args_accepts_direct_mode_and_global_overrides():
+    args = parse_args([
+        "llm",
+        "--url", "http://127.0.0.1:6745",
+        "--token", "secret-token",
+        "--model", "text:test-model",
+        "--prompt", "Say hello",
+        "--output-dir", "tmp/out",
+    ])
+
+    assert args.mode == "llm"
+    assert args.url == "http://127.0.0.1:6745"
+    assert args.token == "secret-token"
+    assert args.model == "text:test-model"
+    assert args.prompt == "Say hello"
+    assert args.output_dir == "tmp/out"
+
+
+def test_choose_mode_interactively_maps_numeric_selection(monkeypatch):
+    monkeypatch.setattr("builtins.input", lambda _: "2")
+
+    mode = choose_mode_interactively()
+
+    assert mode == "transcription"
+
+
+def test_parse_args_leaves_mode_empty_for_interactive_fallback():
+    args = parse_args([])
+
+    assert args.mode is None

tools/manual_multimodal_test_client.py

+from __future__ import annotations
+
+import argparse
+from pathlib import Path
+from typing import Optional
+
+
+MODES = [
+    "llm",
+    "transcription",
+    "audio-generation",
+    "video-generation",
+    "video-doubt",
+    "music-audio-doubt",
+]
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(description="Manual multimodal smoke-test client")
+    parser.add_argument("mode", nargs="?", choices=MODES)
+    parser.add_argument("--url", default="http://127.0.0.1:6745")
+    parser.add_argument("--token", default=None)
+    parser.add_argument("--model", default=None)
+    parser.add_argument("--prompt", default=None)
+    parser.add_argument("--output-dir", default="tmp/manual-client-output")
+    parser.add_argument("--file", default=None)
+    parser.add_argument("--audio-file", default=None)
+    parser.add_argument("--video-file", default=None)
+    return parser
+
+
+def parse_args(argv: list[str] | None = None) -> argparse.Namespace:
+    return build_parser().parse_args(argv)
+
+
+def choose_mode_interactively() -> str:
+    for idx, mode in enumerate(MODES, start=1):
+        print(f"{idx}. {mode}")
+    raw = input("Choose mode: ").strip()
+    selected = int(raw)
+    if selected < 1 or selected > len(MODES):
+        raise ValueError(f"Invalid mode selection: {raw}")
+    return MODES[selected - 1]