feat: inline markdown output box — render AI reply with marked, do not replace cell

Flow (v3-final corrected):
  1. Model call succeeds.
  2. Server passes the AI reply through unchanged as 'markdown' (the
     system prompt allows ```language fences + a brief explanation, so
     the response is real markdown that marked can render into code
     blocks, headings, etc.).
  3. Client OpenCodeCellActions._handleResponse calls prompt.setOutput(
     resp.markdown); the inline prompt widget renders it with
     marked.parse and shows it in a new output area (hidden until a
     response arrives). The cell source is NOT replaced.
  4. User can close the output or cancel the whole prompt.

Server:
  - New unified system prompt: '你是代码助手 ... 按指令修改代码，可附简
    短说明' (allows ```fences``` + explanation; no more 'no markdown
    fences' restriction).
  - EditHandler returns {ok, markdown, sessionId, notebookPath} (raw
    text, fences intact). _strip_code_fence kept as a helper for any
    future apply-to-cell path; no longer called.
  - finalSource field dropped (the cell-apply path is gone).

Client:
  - OpenCodeSuccess: markdown: string (finalSource removed).
  - OpenCodeInlinePrompt: new output area, setOutput(md) renders via
    marked.parse, hideOutput() closes it.
  - OpenCodeCellActions._handleResponse: setOutput(markdown) instead of
    sharedModel.setSource + auto-hide. The prompt stays open so the
    user can read the output.
  - Uses marked@17 (already in node_modules via JupyterLab; no new dep).
  - CSS: output area styling (border, max-height 320px scroll, code/pre
    styling, close button).

Tests: pytest 34/34, jest 29/29. marked is mocked in jest.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

opencode_bridge/tests/test_routes.py

@@ -135,7 +135,9 @@ async def test_edit_handler(monkeypatch, jp_fetch):
     assert response.code == 200
     payload = json.loads(response.body)
     assert payload["ok"] is True
-    assert payload["finalSource"] == "def foo():\n    return 42"
+    # Server now returns the AI reply as raw markdown (```fences``` kept
+    # so the frontend marked.parse renders code blocks).
+    assert payload["markdown"] == "def foo():\n    return 42"
     assert payload["sessionId"] == "fake-session-123"
     assert payload["notebookPath"] == "test.ipynb"
 
@@ -148,8 +150,8 @@ async def test_edit_handler(monkeypatch, jp_fetch):
     # send_message_sync received the session ID from the manager
     send_call = [c for c in fake.calls if c[0] == "send_message_sync"][0]
     assert send_call[1] == "fake-session-123"
-    # system prompt was passed
-    assert "你是一个代码编辑助手" in send_call[3]
+    # system prompt was passed (v3+ allows markdown/fences in the reply)
+    assert "你是一个代码助手" in send_call[3]
 
     # SessionManager.get_or_create was called with the notebook path
     sm_call_names = [c[0] for c in fake_sm.calls]