Replace temperature with reasoning effort and retry in-stream errors

- Providers get a default reasoning effort (low..max, empty = not sent);
  chat can override it per message. OpenAI sends reasoning_effort,
  Claude sends output_config.effort with adaptive thinking, and thinking
  blocks are replayed during the tool loop.
- Show a thinking indicator while waiting for model output.
- Surface in-stream upstream errors (e.g. overloaded) instead of
  breaking the SSE stream, and retry them up to 3 times when nothing
  has been emitted yet.

backend/alembic/versions/20261014_5c2e8f1a7b3d_provider_reasoning_effort.py

@@ -0,0 +1,31 @@
+"""provider reasoning effort replaces temperature
+
+Revision ID: 5c2e8f1a7b3d
+Revises: 8349ae014ddd
+Create Date: 2026-10-14 14:00:00.000000
+
+"""
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision: str = '5c2e8f1a7b3d'
+down_revision: Union[str, Sequence[str], None] = '8349ae014ddd'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    """Upgrade schema."""
+    # 用户决定不再设置 temperature（2026-10-14），改为思考强度
+    op.drop_column('ai_providers', 'temperature')
+    op.add_column('ai_providers', sa.Column('reasoning_effort', sa.String(length=8), nullable=True))
+
+
+def downgrade() -> None:
+    """Downgrade schema."""
+    op.drop_column('ai_providers', 'reasoning_effort')
+    op.add_column('ai_providers', sa.Column('temperature', sa.Float(), nullable=True))