feat: add Chinese UI, Responses models, and mirrored Windows setup

Localize the web console and persist custom model and capability settings.
Use OpenAI Responses exclusively and move runtime limits into environment configuration.
Bootstrap uv and the virtual environment on Windows, support optional China mirrors,
and package source distributions without environments, build output, or credentials.

Update regression tests and repository quality checks for the new behavior.

tests/unit/agents/test_operator_transcript.py

@@ -54,8 +54,8 @@ from artemis.memory.transcript import (
 
 # SHA-256 snapshots of the legacy system message with the fixed inputs below.
 # Update these when an intentional template change alters the rendered prompt.
-GOLDEN_EMPTY_PLAN = "513e0341014b9a2342f608ad4dd173b7dcf5704ce53ee028ba302a9a1b696f96"
-GOLDEN_SENTINEL_PLAN = "012fe1b08645266d5a7e518023e20fc6fbfa65d4f2d985847ebb06e0184c67ee"
+GOLDEN_EMPTY_PLAN = "d715128625911aec46c83138e15aabfa967b385ab4c878509fe36ab4304fae2a"
+GOLDEN_SENTINEL_PLAN = "e2cf4ede75ffd9e283159f70dcd5c94a169c18e9f6dd4d1e7aacf460e956eaf3"
 
 SCREENSHOT_B64 = base64.b64encode(b"fake-jpeg-bytes").decode("utf-8")
 
@@ -77,7 +77,7 @@ async def _render_legacy_system(plan_and_history: str) -> str:
 
 
 @pytest.mark.asyncio
-async def test_legacy_prompt_bytes_match_pre_split_golden():
+async def test_legacy_prompt_bytes_match_chinese_prompt_golden():
     empty = await _render_legacy_system("No plan or history yet.")
     sentinel = await _render_legacy_system("PLAN_AND_HISTORY_SENTINEL\nline2")
     assert hashlib.sha256(empty.encode("utf-8")).hexdigest() == GOLDEN_EMPTY_PLAN