feat(user): 昵称上限放宽到 20 字(与客户端/原型一致)

ProfileUpdateRequest.nickname 的 max_length 16 → 20,与客户端 take(20) 和
原型 settings.html(maxlength=20)对齐。修复 17–20 字昵称保存时后端返回 422、
客户端弹「更新失败,请重试」的问题。DB 列为 String(64),放得下,无需迁移。

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

.env.example

@@ -83,6 +83,8 @@ WXPAY_AUTH_NOTIFY_URL=
 WITHDRAW_AUTO_RECONCILE_ENABLED=false
 WITHDRAW_AUTO_RECONCILE_INTERVAL_SEC=300
 WITHDRAW_AUTO_RECONCILE_OLDER_THAN_MINUTES=15
+# 0 点自动兑金币开关(deploy/daily-exchange.timer 触发的脚本读它;false=脚本 no-op)。默认 true。
+AUTO_EXCHANGE_ENABLED=true
 
 # ===== 穿山甲激励视频(服务端发奖回调)=====
 # 看完激励视频后穿山甲服务器 S2S 回调本服务发金币(客户端不参与发奖)。

alembic/versions/drop_user_force_onboarding.py

@@ -0,0 +1,40 @@
+"""drop user.force_onboarding column
+
+force_onboarding(运营「按用户」强制新手引导)机制废弃:客户端实际按 onboarding_completion
+(设备+账号 维度)判断是否走引导,该 force_onboarding 列从未被客户端正确响应(被
+onboarding_completed 压过)。改用 admin「设备维度」重置(删 onboarding_completion 记录)替代,
+见 app/admin/routers/onboarding.py。本迁移删列。
+
+Revision ID: drop_force_onboarding
+Revises: 9b894f5fff05
+Create Date: 2026-06-12 00:00:00.000000
+
+"""
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision: str = "drop_force_onboarding"
+down_revision: Union[str, Sequence[str], None] = "9b894f5fff05"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    op.drop_column("user", "force_onboarding")
+
+
+def downgrade() -> None:
+    # 回滚:重建列(同 user_force_onboarding 迁移的 upgrade)。sa.false() 两端兼容。
+    op.add_column(
+        "user",
+        sa.Column(
+            "force_onboarding",
+            sa.Boolean(),
+            nullable=False,
+            server_default=sa.false(),
+        ),
+    )

app/admin/main.py

@@ -21,6 +21,7 @@ from app.admin.routers.config import router as config_router
 from app.admin.routers.dashboard import router as dashboard_router
 from app.admin.routers.ops_stat_config import router as ops_stat_config_router
 from app.admin.routers.feedback import router as feedback_router
+from app.admin.routers.onboarding import router as onboarding_router
 from app.admin.routers.ops_marquee_seed import router as ops_marquee_seed_router
 from app.admin.routers.price_report import router as price_report_router
 from app.admin.routers.users import router as users_router
@@ -80,6 +81,7 @@ admin_app.include_router(dashboard_router)
 admin_app.include_router(ops_stat_config_router)
 admin_app.include_router(ops_marquee_seed_router)
 admin_app.include_router(users_router)
+admin_app.include_router(onboarding_router)
 admin_app.include_router(wallet_router)
 admin_app.include_router(withdraw_router)
 admin_app.include_router(price_report_router)

app/admin/repositories/mutations.py

@@ -10,10 +10,12 @@ from __future__ import annotations
 
 from datetime import datetime
 
+from sqlalchemy import delete
 from sqlalchemy.orm import Session
 
 from app.core.rewards import CN_TZ
 from app.models.feedback import Feedback
+from app.models.onboarding import OnboardingCompletion
 from app.models.price_report import PriceReport
 from app.models.user import User
 
@@ -42,21 +44,6 @@ def set_user_debug_trace(
     return user
 
 
-def set_user_force_onboarding(
-    db: Session, user: User, *, enabled: bool, commit: bool = True
-) -> User:
-    """运营「一键开启/取消新手引导」:置 user.force_onboarding。开启后该用户下次启动 App 被强制
-    重走引导教程,走完即由 /onboarding/complete 自动清回 false。同 set_user_debug_trace:支持
-    commit=False 让 router 把业务写 + 审计写放进同一事务。"""
-    user.force_onboarding = enabled
-    if commit:
-        db.commit()
-        db.refresh(user)
-    else:
-        db.flush()
-    return user
-
-
 def update_feedback_status(
     db: Session, feedback: Feedback, *, status: str, commit: bool = True
 ) -> Feedback:
@@ -94,3 +81,27 @@ def review_price_report(
     else:
         db.flush()
     return report
+
+
+def delete_device_onboarding(db: Session, device_id: str, *, commit: bool = True) -> int:
+    """删某设备(device_id)的全部 onboarding 完成记录 → 该设备上所有账号下次登录重走引导。
+    返回删除行数。支持 commit=False 让 router 把业务写 + 审计写放进同一事务。"""
+    n = db.execute(
+        delete(OnboardingCompletion).where(OnboardingCompletion.device_id == device_id)
+    ).rowcount
+    if commit:
+        db.commit()
+    else:
+        db.flush()
+    return n
+
+
+def delete_all_onboarding(db: Session, *, commit: bool = True) -> int:
+    """清空 onboarding_completion 表 → 所有用户在所有设备下次登录都重走引导。返回删除行数。
+    支持 commit=False(同 delete_device_onboarding)。"""
+    n = db.execute(delete(OnboardingCompletion)).rowcount
+    if commit:
+        db.commit()
+    else:
+        db.flush()
+    return n

app/admin/repositories/queries.py

@@ -14,6 +14,7 @@ from sqlalchemy.orm import Session
 from app.models.admin import AdminAuditLog
 from app.models.comparison import ComparisonRecord
 from app.models.feedback import Feedback
+from app.models.onboarding import OnboardingCompletion
 from app.models.price_report import PriceReport
 from app.models.user import User
 from app.models.wallet import CashTransaction, CoinAccount, CoinTransaction, WithdrawOrder
@@ -56,6 +57,26 @@ def list_users(
     return cursor_paginate(db, stmt, User.id, limit=limit, cursor=cursor)
 
 
+def list_onboarding_devices(db: Session, *, limit: int = 500) -> list[dict]:
+    """按设备(device_id, ANDROID_ID)聚合 onboarding_completion:每台设备走过引导的账号数 +
+    最近完成时间,按最近完成倒序。设备维度新手引导管理用。没走过引导的设备不在表里(本就会
+    引导、无需管理)。当前全量返回(上限 limit;调试期设备少,量大再加分页/搜索)。"""
+    rows = db.execute(
+        select(
+            OnboardingCompletion.device_id,
+            func.count(OnboardingCompletion.id),
+            func.max(OnboardingCompletion.completed_at),
+        )
+        .group_by(OnboardingCompletion.device_id)
+        .order_by(func.max(OnboardingCompletion.completed_at).desc())
+        .limit(limit)
+    ).all()
+    return [
+        {"device_id": did, "account_count": int(cnt), "last_completed_at": last}
+        for did, cnt, last in rows
+    ]
+
+
 def list_all_coin_transactions(
     db: Session,
     *,

app/admin/routers/onboarding.py

@@ -0,0 +1,86 @@
+"""admin 设备维度新手引导管理:列设备 / 重置单设备 / 全部重置。
+
+背景:原 force_onboarding(运营「按用户」强制引导)客户端不认(被 onboarding_completed 压过),
+已废弃。改为直接操作 onboarding_completion(设备+账号 维度):删记录 → 该(设备,账号)的
+onboarding_completed 变 false → 客户端任何版本下次登录都重走引导。可靠、不依赖客户端实现。
+"""
+from __future__ import annotations
+
+from typing import Annotated
+
+from fastapi import APIRouter, Depends, HTTPException, Request
+
+from app.admin.audit import write_audit
+from app.admin.deps import AdminDb, get_client_ip, get_current_admin, require_role
+from app.admin.repositories import mutations, queries
+from app.admin.schemas.common import CursorPage, OkResponse
+from app.admin.schemas.onboarding import DeviceOnboardingItem, ResetAllRequest
+from app.models.admin import AdminUser
+
+router = APIRouter(
+    prefix="/admin/api/onboarding",
+    tags=["admin-onboarding"],
+    dependencies=[Depends(get_current_admin)],
+)
+
+
+@router.get(
+    "/devices",
+    response_model=CursorPage[DeviceOnboardingItem],
+    summary="设备列表(按设备聚合走过引导的账号)",
+)
+def list_devices(db: AdminDb) -> CursorPage[DeviceOnboardingItem]:
+    """列出走过新手引导的设备(ANDROID_ID),按最近完成时间倒序。没走过引导的设备不在表里、
+    本就会引导,无需管理,故不列。当前全量返回(上限 500;设备量大再加分页/搜索),
+    next_cursor 恒 None。"""
+    rows = queries.list_onboarding_devices(db)
+    return CursorPage(
+        items=[DeviceOnboardingItem(**r) for r in rows],
+        next_cursor=None,
+    )
+
+
+@router.post(
+    "/devices/{device_id}/reset",
+    response_model=OkResponse,
+    summary="重置单个设备(该设备所有账号下次重走引导)",
+)
+def reset_device(
+    device_id: str,
+    request: Request,
+    admin: Annotated[AdminUser, Depends(require_role("operator"))],
+    db: AdminDb,
+) -> OkResponse:
+    """删该设备的全部 onboarding 完成记录 → 该设备上所有账号下次登录重走引导。"""
+    # 业务写 + 审计写同一事务(commit=False),最后一起 commit(同 users 各写接口)
+    deleted = mutations.delete_device_onboarding(db, device_id, commit=False)
+    write_audit(
+        db, admin, action="device_onboarding.reset", target_type="device", target_id=device_id,
+        detail={"deleted_rows": deleted}, ip=get_client_ip(request), commit=False,
+    )
+    db.commit()
+    return OkResponse()
+
+
+@router.post(
+    "/reset-all",
+    response_model=OkResponse,
+    summary="全部重设(清空所有设备引导记录)",
+)
+def reset_all(
+    body: ResetAllRequest,
+    request: Request,
+    admin: Annotated[AdminUser, Depends(require_role("operator"))],
+    db: AdminDb,
+) -> OkResponse:
+    """清空 onboarding_completion 表 → 所有用户在所有设备下次登录都重走引导。
+    需 confirm=true 防误触(前端另有二次确认弹窗)。"""
+    if not body.confirm:
+        raise HTTPException(status_code=400, detail="需 confirm=true 确认")
+    deleted = mutations.delete_all_onboarding(db, commit=False)
+    write_audit(
+        db, admin, action="device_onboarding.reset_all", target_type="onboarding", target_id=None,
+        detail={"deleted_rows": deleted}, ip=get_client_ip(request), commit=False,
+    )
+    db.commit()
+    return OkResponse()

app/admin/routers/users.py

@@ -15,7 +15,6 @@ from app.admin.schemas.user import (
     GrantCashRequest,
     GrantCoinsRequest,
     SetDebugTraceRequest,
-    SetForceOnboardingRequest,
     SetUserStatusRequest,
 )
 from app.models.admin import AdminUser
@@ -102,33 +101,6 @@ def set_user_debug_trace(
     return OkResponse()
 
 
-@router.post("/{user_id}/force-onboarding", response_model=OkResponse, summary="一键开启/取消新手引导")
-def set_user_force_onboarding(
-    user_id: int,
-    body: SetForceOnboardingRequest,
-    request: Request,
-    admin: Annotated[AdminUser, Depends(require_role("operator"))],
-    db: AdminDb,
-) -> OkResponse:
-    """运营「一键开启新手引导」:置 user.force_onboarding。开启后该用户下次启动 App(/me 带出此字段)
-    会被强制重走引导教程——即便此前已看完;走完引导后端自动清回 false,不无限循环。
-    仅对支持 force_onboarding 的 App 版本生效(老版本忽略该字段)。"""
-    user = user_repo.get_user_by_id(db, user_id)
-    if user is None:
-        raise HTTPException(status_code=404, detail="用户不存在")
-    if user.status == "deleted":
-        raise HTTPException(status_code=400, detail="已注销账号不可操作")
-    before = user.force_onboarding
-    # 业务写 + 审计写同一事务(commit=False),最后一起 commit(同 set_user_status / debug-trace)
-    mutations.set_user_force_onboarding(db, user, enabled=body.enabled, commit=False)
-    write_audit(
-        db, admin, action="user.force_onboarding.set", target_type="user", target_id=user_id,
-        detail={"before": before, "after": body.enabled}, ip=get_client_ip(request), commit=False,
-    )
-    db.commit()
-    return OkResponse()
-
-
 @router.post("/{user_id}/coins", response_model=OkResponse, summary="手动增减金币(带审计)")
 def grant_user_coins(
     user_id: int,

app/admin/schemas/onboarding.py

@@ -0,0 +1,24 @@
+"""admin 设备维度新手引导管理 schemas。
+
+force_onboarding(按用户)废弃后改用「按设备」:直接操作 onboarding_completion(设备+账号 维度),
+客户端任何版本都按它判断是否走引导,故可靠。
+"""
+from __future__ import annotations
+
+from datetime import datetime
+
+from pydantic import BaseModel, Field
+
+
+class DeviceOnboardingItem(BaseModel):
+    """一台设备(ANDROID_ID)的引导聚合:在该设备走过引导的账号数 + 最近完成时间。"""
+
+    device_id: str
+    account_count: int
+    last_completed_at: datetime
+
+
+class ResetAllRequest(BaseModel):
+    confirm: bool = Field(
+        False, description="必须 true 才清空全部设备的引导记录(防误调接口;前端另有二次确认弹窗)"
+    )

app/admin/schemas/user.py

@@ -16,8 +16,6 @@ class AdminUserListItem(BaseModel):
     register_channel: str
     status: str
     debug_trace_enabled: bool = False
-    # 运营「一键开启新手引导」:true=该用户下次启动 App 被强制重走引导(走完自动清回 false)
-    force_onboarding: bool = False
     wechat_openid: str | None = None
     created_at: datetime
     last_login_at: datetime
@@ -57,9 +55,3 @@ class SetUserStatusRequest(BaseModel):
 
 class SetDebugTraceRequest(BaseModel):
     enabled: bool = Field(..., description="是否给该用户开「复制调试链接」权限")
-
-
-class SetForceOnboardingRequest(BaseModel):
-    enabled: bool = Field(
-        ..., description="true=一键开启(强制该用户下次启动 App 重走新手引导)/ false=取消"
-    )

app/api/v1/user.py

@@ -12,14 +12,19 @@ from __future__ import annotations
 
 import logging
 
-from fastapi import APIRouter, File, HTTPException, UploadFile
+from fastapi import APIRouter, File, HTTPException, Query, UploadFile
 
 from app.api.deps import CurrentUser, DbSession
 from app.core import media
 from app.repositories import onboarding as onboarding_repo
 from app.repositories import user as user_repo
 from app.schemas.auth import UserOut
-from app.schemas.user import OkResponse, OnboardingCompleteRequest, ProfileUpdateRequest
+from app.schemas.user import (
+    OkResponse,
+    OnboardingCompleteRequest,
+    OnboardingStatusResponse,
+    ProfileUpdateRequest,
+)
 
 logger = logging.getLogger("shagua.user")
 
@@ -59,12 +64,27 @@ def complete_onboarding(
     """走完新手引导时调一次。按 (当前账号, device_id) 落一条完成标记,跨卸载重装持久。
     幂等:重复调用不报错。device_id 取客户端硬件级 ANDROID_ID,与登录请求一致。"""
     onboarding_repo.mark_completed(db, user_id=user.id, device_id=req.device_id)
-    # 若该用户被运营「一键开启新手引导」强制拉回引导,走完即清除强制标记(否则下次启动还会再触发)。
-    user_repo.clear_force_onboarding(db, user)
     logger.info("onboarding complete user_id=%d device_len=%d", user.id, len(req.device_id))
     return OkResponse()
 
 
+@router.get(
+    "/onboarding/status",
+    response_model=OnboardingStatusResponse,
+    summary="查新手引导是否已完成(按 设备+账号)",
+)
+def onboarding_status(
+    user: CurrentUser,
+    db: DbSession,
+    device_id: str = Query("", max_length=64, description="硬件级 ANDROID_ID;空=按未完成处理"),
+) -> OnboardingStatusResponse:
+    """已登录用户启动时查:该 (账号, 设备) 走过引导没。false → 客户端重走(运营在 admin 删了该设备
+    记录即触发)。替代原 force_onboarding(按用户)——改设备维度后这是"运营触发重走"的查询入口。
+    device_id 为空一律按未完成(同登录:不误跳过)。"""
+    completed = onboarding_repo.is_completed(db, user_id=user.id, device_id=device_id)
+    return OnboardingStatusResponse(completed=completed)
+
+
 @router.delete("", response_model=OkResponse, summary="注销账号(软删除)")
 def delete_account(user: CurrentUser, db: DbSession) -> OkResponse:
     media.delete_avatar(user.avatar_url)

app/core/config.py

@@ -105,6 +105,10 @@ class Settings(BaseSettings):
     WITHDRAW_AUTO_RECONCILE_ENABLED: bool = False
     WITHDRAW_AUTO_RECONCILE_INTERVAL_SEC: int = 300
     WITHDRAW_AUTO_RECONCILE_OLDER_THAN_MINUTES: int = 15
+    # 0 点自动兑金币(由 deploy/daily-exchange.timer 触发 scripts.daily_auto_exchange)。
+    # 客户端已删手动兑入口、改为「0 点自动兑现金」,故默认开;运营要临时停可在 .env 置 false,
+    # 无需动 systemd timer——脚本读此开关,false 时直接 no-op 退出。
+    AUTO_EXCHANGE_ENABLED: bool = True
     # 免确认收款授权(用户授权免确认模式)的授权结果回调地址,必须公网可访问 HTTPS、不带参数。
     # 发起授权 / 首单顺带授权时作为 authorization_notify_url 传给微信。一期不处理回调内容
     # (授权状态靠 query 查询兜底),但微信要求该字段非空,故启用免确认前必须配置;留空时免确认相关接口返回未配置。

app/models/user.py

@@ -58,13 +58,6 @@ class User(Base):
         Boolean, nullable=False, default=False, server_default=false()
     )
 
-    # 运营后台「一键开启新手引导」:置 true 后,该用户下次启动 App(/me 与登录响应带出此字段)会被
-    # 强制重走新手引导教程——即便本地早已标记完成。走完引导(/onboarding/complete)即自动清回 false,
-    # 不会无限循环。默认 false。
-    force_onboarding: Mapped[bool] = mapped_column(
-        Boolean, nullable=False, default=False, server_default=false()
-    )
-
     created_at: Mapped[datetime] = mapped_column(
         DateTime(timezone=True), server_default=func.now(), nullable=False
     )

app/repositories/user.py

@@ -62,16 +62,6 @@ def set_avatar_url(db: Session, user: User, *, avatar_url: str) -> User:
     return user
 
 
-def clear_force_onboarding(db: Session, user: User) -> None:
-    """用户走完(被运营强制开启的)新手引导后,清除强制标记,避免下次启动再被拉回引导。
-
-    幂等:未置位时直接返回,不空 commit。由 /onboarding/complete 在标记完成后调用。
-    """
-    if user.force_onboarding:
-        user.force_onboarding = False
-        db.commit()
-
-
 def soft_delete_account(db: Session, user: User) -> None:
     """注销账号:软删除 + 匿名化。
 

app/repositories/wallet.py

@@ -183,14 +183,22 @@ def list_coin_transactions(
 
 
 def exchange_coins_to_cash(
-    db: Session, user_id: int, coin_amount: int
+    db: Session,
+    user_id: int,
+    coin_amount: int,
+    *,
+    remark: str | None = None,
+    enforce_min: bool = True,
 ) -> tuple[CoinAccount, int]:
     """金币兑现金。返回 (account, 本次兑入的分)。
 
-    校验:金额 >= MIN_EXCHANGE_COIN 且为整分倍数(InvalidExchangeAmountError),
-    余额充足(InsufficientCoinError)。扣金币 + 加现金,两条流水同事务 commit。
+    校验:金额为整分倍数(InvalidExchangeAmountError)、余额充足(InsufficientCoinError);
+    `enforce_min=True`(手动兑默认)时还要 >= 后台配置的兑换下限,`enforce_min=False`
+    (0 点自动兑「到分全额」用)只要够 1 分(COIN_PER_CENT)即可,不受手动下限约束。
+    `remark` 不传则用默认「N金币兑M分」。扣金币 + 加现金,两条流水同事务 commit。
     """
-    if coin_amount < rewards.get_min_exchange_coin(db) or coin_amount % COIN_PER_CENT != 0:
+    floor_min = rewards.get_min_exchange_coin(db) if enforce_min else COIN_PER_CENT
+    if coin_amount < floor_min or coin_amount % COIN_PER_CENT != 0:
         raise InvalidExchangeAmountError
 
     acc = get_or_create_account(db, user_id, commit=False)
@@ -198,10 +206,10 @@ def exchange_coins_to_cash(
         raise InsufficientCoinError
 
     cents = coins_to_cents(coin_amount)
-    remark = f"{coin_amount}金币兑{cents}分"
+    tx_remark = remark or f"{coin_amount}金币兑{cents}分"
 
     # 1. 扣金币(写 coin_transaction)
-    grant_coins(db, user_id, -coin_amount, biz_type="exchange_out", remark=remark)
+    grant_coins(db, user_id, -coin_amount, biz_type="exchange_out", remark=tx_remark)
     # 2. 加现金(写 cash_transaction)
     acc.cash_balance_cents += cents
     db.add(
@@ -210,7 +218,7 @@ def exchange_coins_to_cash(
             amount_cents=cents,
             balance_after_cents=acc.cash_balance_cents,
             biz_type="exchange_in",
-            remark=remark,
+            remark=tx_remark,
             created_at=datetime.now(rewards.CN_TZ).replace(tzinfo=None),
         )
     )
@@ -219,6 +227,70 @@ def exchange_coins_to_cash(
     return acc, cents
 
 
+def _has_exchange_in_on(db: Session, user_id: int, day) -> bool:
+    """该用户在指定日期(北京日)是否已有 exchange_in 现金流水。
+
+    自动兑复用 exchange_in 类型,故用「当天是否已有 exchange_in」做幂等键:timer 重触发 /
+    手动重跑当天不会重复兑。手动兑入口已下线,不会与之误判混淆(若日后恢复手动兑需改用独立类型)。
+    created_at 落库为北京 naive 时间(见 exchange_coins_to_cash),按 [当天 0 点, 次日 0 点) 比较。
+    """
+    day_start = datetime.combine(day, datetime.min.time())
+    next_day = day_start + timedelta(days=1)
+    return (
+        db.execute(
+            select(CashTransaction.id)
+            .where(
+                CashTransaction.user_id == user_id,
+                CashTransaction.biz_type == "exchange_in",
+                CashTransaction.created_at >= day_start,
+                CashTransaction.created_at < next_day,
+            )
+            .limit(1)
+        ).first()
+        is not None
+    )
+
+
+def daily_auto_exchange(db: Session) -> dict:
+    """0 点定时任务:把每个用户金币「到分全额」兑成现金(复用 exchange_in 流水类型)。
+
+    - **到分全额**:coin_amount = floor(余额 / 100) * 100;余额不足 1 分(<100)的零头留到下次。
+    - **幂等**:当天已有 exchange_in 的用户跳过(见 [_has_exchange_in_on])。
+    - **逐用户独立事务**:单个用户异常 rollback 不影响其他人;exchange_coins_to_cash 内部按用户 commit。
+    返回统计 dict(scanned/converted/skipped_done/skipped_dust/failed/total_cents)。
+    """
+    today = rewards.cn_today()
+    stats = {
+        "scanned": 0, "converted": 0, "skipped_done": 0,
+        "skipped_dust": 0, "failed": 0, "total_cents": 0,
+    }
+    user_ids = db.execute(
+        select(CoinAccount.user_id).where(CoinAccount.coin_balance >= COIN_PER_CENT)
+    ).scalars().all()
+
+    for user_id in user_ids:
+        stats["scanned"] += 1
+        try:
+            if _has_exchange_in_on(db, user_id, today):
+                stats["skipped_done"] += 1
+                continue
+            acc = get_or_create_account(db, user_id, commit=False)
+            coin_amount = (acc.coin_balance // COIN_PER_CENT) * COIN_PER_CENT
+            if coin_amount < COIN_PER_CENT:
+                stats["skipped_dust"] += 1
+                continue
+            _, cents = exchange_coins_to_cash(
+                db, user_id, coin_amount, remark="0 点自动兑现金", enforce_min=False
+            )
+            stats["converted"] += 1
+            stats["total_cents"] += cents
+        except Exception:  # noqa: BLE001 - 批处理不因单用户异常中断
+            db.rollback()
+            stats["failed"] += 1
+
+    return stats
+
+
 def list_cash_transactions(
     db: Session,
     user_id: int,

app/schemas/auth.py

@@ -27,9 +27,6 @@ class UserOut(BaseModel):
     last_login_at: datetime
     # 调试链接权限:前端据此在比价结果弹窗/记录页显示「复制调试链接」按钮。默认 false。
     debug_trace_enabled: bool = False
-    # 运营后台「一键开启新手引导」:true → 客户端下次启动强制重走引导(即便本地已完成);
-    # 走完引导后端自动清回 false。默认 false 兼容老客户端。
-    force_onboarding: bool = False
 
 
 # ===== Token 通用结构 =====

app/schemas/user.py

@@ -5,7 +5,7 @@ from pydantic import BaseModel, Field, field_validator
 
 
 class ProfileUpdateRequest(BaseModel):
-    nickname: str = Field(..., min_length=1, max_length=16, description="昵称,1-16 字")
+    nickname: str = Field(..., min_length=1, max_length=20, description="昵称,1-20 字")
 
     @field_validator("nickname")
     @classmethod
@@ -23,5 +23,9 @@ class OnboardingCompleteRequest(BaseModel):
     )
 
 
+class OnboardingStatusResponse(BaseModel):
+    completed: bool = Field(..., description="该 设备+账号 是否已走过新手引导(false → 客户端应重走)")
+
+
 class OkResponse(BaseModel):
     ok: bool = True

deploy/daily-exchange.md

@@ -0,0 +1,59 @@
+# 0 点自动兑金币 定时任务 — 运维手册
+
+> 对象:维护「每天 0 点把用户金币自动兑成现金」这套定时任务的同事。
+> 🔒 服务器登录信息见**私密交接清单**,不入库。
+
+## 它是什么
+客户端已删手动「兑金币」入口、改为「0 点自动兑现金（可能存在延迟）」,故服务端每天 0 点跑一轮:
+把每个用户的金币按汇率 **100 金币 = 1 分** 兑成现金。
+
+- **到分全额**:`coin_amount = floor(余额 / 100) × 100`,不足 1 分(<100 金币)的零头留到下次。
+- **复用 `exchange_in` 流水类型**(手动兑入口已下线,不会混淆),用户在「现金明细」看到「金币兑换现金」。
+- **幂等**:同一用户当天已有 `exchange_in` 流水则跳过 → timer 重触发 / 手动重跑当天不会重复兑。
+- 逻辑见 `app/repositories/wallet.py::daily_auto_exchange`,入口 `scripts/daily_auto_exchange.py`。
+
+## 文件
+| 项 | 路径 |
+|---|---|
+| 脚本入口 | `scripts/daily_auto_exchange.py` |
+| 核心逻辑 | `app/repositories/wallet.py::daily_auto_exchange` |
+| 开关 | `.env` 的 `AUTO_EXCHANGE_ENABLED`(默认 true;false=脚本 no-op) |
+| systemd 单元 | `deploy/daily-exchange.{service,timer}` |
+| 运行锁 | `/tmp/daily_auto_exchange.lock`(可用环境变量 `DAILY_EXCHANGE_LOCK` 覆盖) |
+
+## 部署(Linux 服务器)
+```bash
+sudo cp deploy/daily-exchange.{service,timer} /etc/systemd/system/
+sudo systemctl daemon-reload && sudo systemctl enable --now daily-exchange.timer
+systemctl list-timers daily-exchange.timer   # 确认下次触发时间
+```
+
+## 本机 Windows 开发(无 systemd)
+直接手动跑:
+```
+.venv\Scripts\python -m scripts.daily_auto_exchange --once
+# 本地循环调试(每小时一轮,幂等空跑无害):
+.venv\Scripts\python -m scripts.daily_auto_exchange --loop --interval 3600
+```
+
+## 怎么看健康 / 手动跑一次
+```bash
+sudo systemctl start daily-exchange.service     # 立即手动跑一轮(不等 0 点)
+journalctl -u daily-exchange -n 30 --no-pager   # 看日志:扫描/兑换/跳过/失败/累计兑入分
+```
+日志一行形如:`自动兑完成 用时1.2s 扫描120 兑换88 已兑过跳过30 零头跳过2 失败0 累计兑入15600分`。
+
+## 怎么关 / 临时停
+- **临时停(推荐)**:`.env` 置 `AUTO_EXCHANGE_ENABLED=false` → 脚本 no-op,无需动 timer。
+- **停 timer**:`sudo systemctl disable --now daily-exchange.timer`。
+
+## 脚本参数
+- `--once`:跑一轮(timer 用这个,默认)
+- `--loop --interval N`:常驻循环,每 N 秒一轮(默认 3600,调试用)
+- 锁位置可用环境变量 `DAILY_EXCHANGE_LOCK=/path/to.lock` 覆盖
+
+## 注意事项
+- **触发时间**:`OnCalendar=*-*-* 00:00:00`。客户端文案已注明「可能存在延迟」,要错开整点扎堆可改 `00:05:00`。
+- **catch-up**:`Persistent=true`,服务器宕机错过当天会在重启后补跑一轮。
+- **DB 无关**:sqlite / postgres 均可(不像美团 ETL 需要 PG)。
+- **改脚本 / 改部署**:走 git + PR,由有 root 的人部署。

deploy/daily-exchange.service

@@ -0,0 +1,33 @@
+# 0 点自动兑金币 —— 单轮把每个用户金币「到分全额」兑成现金,由 daily-exchange.timer 每天 0 点触发。
+#
+# 仅用于 Linux 服务器;本机 Windows 开发无 systemd,直接手动跑脚本即可:
+#   .venv\Scripts\python -m scripts.daily_auto_exchange --once
+#
+# 部署(服务器):
+#   sudo cp deploy/daily-exchange.{service,timer} /etc/systemd/system/
+#   sudo systemctl daemon-reload && sudo systemctl enable --now daily-exchange.timer
+#   # 手动跑一次验证: sudo systemctl start daily-exchange.service && journalctl -u daily-exchange -n 30
+#
+# 前置:.env 的 AUTO_EXCHANGE_ENABLED=true(默认);置 false 则脚本 no-op,免动 timer。
+[Unit]
+Description=Daily auto-exchange coins to cash (one-shot, driven by timer)
+After=network-online.target
+Wants=network-online.target
+
+[Service]
+Type=oneshot
+User=root
+WorkingDirectory=/opt/shaguabijia-app-server
+Environment="PATH=/opt/shaguabijia-app-server/.venv/bin:/usr/bin:/bin"
+EnvironmentFile=/opt/shaguabijia-app-server/.env
+ExecStart=/opt/shaguabijia-app-server/.venv/bin/python -m scripts.daily_auto_exchange --once
+SyslogIdentifier=daily-exchange
+# 脚本自带 30min 文件锁;给 20min 硬超时,防卡死轮次长期占锁。
+TimeoutStartSec=1200
+
+# 与主服务 shaguabijia-app-server.service 同款加固。
+NoNewPrivileges=true
+PrivateTmp=true
+ProtectSystem=strict
+ReadWritePaths=/opt/shaguabijia-app-server
+ProtectHome=true

deploy/daily-exchange.timer

@@ -0,0 +1,14 @@
+# 每天 0 点触发一次「自动兑金币」(Linux 服务器用)。
+# 见 daily-exchange.service 顶部注释的部署步骤。
+[Unit]
+Description=Run daily auto-exchange coins->cash at midnight
+
+[Timer]
+# 每天 0 点跑。客户端文案已注明「可能存在延迟」,可按需改 00:05 错开整点扎堆。
+OnCalendar=*-*-* 00:00:00
+# 服务器宕机/重启后,补跑错过的那一轮(而不是干等次日)。
+Persistent=true
+AccuracySec=1min
+
+[Install]
+WantedBy=timers.target

scripts/daily_auto_exchange.py

@@ -0,0 +1,113 @@
+"""0 点自动兑金币(每日定时:把每个用户金币「到分全额」兑成现金)。
+
+背景:客户端已删手动「兑金币」入口、改文案「0 点自动兑现金（可能存在延迟）」,故服务端需在
+每天 0 点把金币按汇率(100 金币 = 1 分)兑成现金。复用 exchange_in 流水类型(手动兑入口已下线,
+不会与之混淆),幂等键 = 用户当天是否已有 exchange_in 流水。核心逻辑见
+app/repositories/wallet.py::daily_auto_exchange。
+
+用法:
+    # 单轮(给 systemd timer / cron 用,线上每天 0 点一次)
+    python -m scripts.daily_auto_exchange --once
+
+    # 本地循环(调试用,默认每小时一轮——真到分兑换幂等,空跑无害)
+    python -m scripts.daily_auto_exchange --loop --interval 3600
+
+开关:.env 的 AUTO_EXCHANGE_ENABLED=false 时脚本直接 no-op 退出(免动 systemd timer)。
+带文件锁,防上一轮没跑完下一轮又起。
+"""
+from __future__ import annotations
+
+import argparse
+import os
+import sys
+import tempfile
+import time
+from datetime import datetime
+
+# Windows 控制台按 UTF-8 输出中文/¥
+try:
+    sys.stdout.reconfigure(encoding="utf-8")  # type: ignore[attr-defined]
+except Exception:  # noqa: BLE001
+    pass
+
+from app.core.config import settings
+from app.db.session import SessionLocal
+from app.repositories import wallet as wallet_repo
+
+# 锁放系统临时目录(任何账号可写),不依赖代码目录 data/ 的写权限。需要指定位置时用环境变量覆盖。
+LOCK_FILE = os.environ.get("DAILY_EXCHANGE_LOCK") or os.path.join(
+    tempfile.gettempdir(), "daily_auto_exchange.lock"
+)
+LOCK_STALE_SEC = 30 * 60   # 锁超过 30min 视为陈旧(进程异常退出残留),自动接管
+
+
+def _acquire_lock() -> bool:
+    os.makedirs(os.path.dirname(LOCK_FILE) or ".", exist_ok=True)
+    try:
+        fd = os.open(LOCK_FILE, os.O_CREAT | os.O_EXCL | os.O_WRONLY)
+        os.write(fd, f"{os.getpid()} {time.time()}".encode())
+        os.close(fd)
+        return True
+    except FileExistsError:
+        try:
+            with open(LOCK_FILE) as f:
+                parts = f.read().split()
+            ts = float(parts[1]) if len(parts) > 1 else 0.0
+            if time.time() - ts > LOCK_STALE_SEC:
+                os.remove(LOCK_FILE)
+                return _acquire_lock()
+        except OSError:
+            pass
+        return False
+
+
+def _release_lock() -> None:
+    try:
+        os.remove(LOCK_FILE)
+    except OSError:
+        pass
+
+
+def run_once() -> None:
+    if not settings.AUTO_EXCHANGE_ENABLED:
+        print(f"[{datetime.now():%H:%M:%S}] AUTO_EXCHANGE_ENABLED=false,跳过(no-op)")
+        return
+    if not _acquire_lock():
+        print(f"[{datetime.now():%H:%M:%S}] 上一轮还在跑(锁占用),跳过本轮")
+        return
+    t0 = time.time()
+    try:
+        with SessionLocal() as db:
+            stats = wallet_repo.daily_auto_exchange(db)
+        dt = time.time() - t0
+        print(
+            f"[{datetime.now():%H:%M:%S}] 自动兑完成 用时{dt:.1f}s "
+            f"扫描{stats['scanned']} 兑换{stats['converted']} "
+            f"已兑过跳过{stats['skipped_done']} 零头跳过{stats['skipped_dust']} "
+            f"失败{stats['failed']} 累计兑入{stats['total_cents']}分"
+        )
+    finally:
+        _release_lock()
+
+
+def main() -> None:
+    ap = argparse.ArgumentParser(description="0 点自动兑金币")
+    ap.add_argument("--once", action="store_true", help="只跑一轮(默认)")
+    ap.add_argument("--loop", action="store_true", help="循环跑(本地调试)")
+    ap.add_argument("--interval", type=int, default=3600, help="循环间隔秒(默认 3600=1h)")
+    args = ap.parse_args()
+
+    if args.loop:
+        print(f"循环模式,每 {args.interval}s 一轮(Ctrl+C 退出)")
+        try:
+            while True:
+                run_once()
+                time.sleep(args.interval)
+        except KeyboardInterrupt:
+            print("已退出")
+    else:
+        run_once()
+
+
+if __name__ == "__main__":
+    main()

tests/test_admin_write.py

@@ -160,48 +160,6 @@ def test_set_user_status_and_audit(admin_client: TestClient, operator_token: str
     ).status_code == 200
 
 
-# ===== 一键开启新手引导 =====
-
-def test_set_force_onboarding_and_audit(admin_client: TestClient, operator_token: str) -> None:
-    uid = _seed_user("13900000009")
-    r = admin_client.post(
-        f"/admin/api/users/{uid}/force-onboarding", json={"enabled": True},
-        headers=_auth(operator_token),
-    )
-    assert r.status_code == 200, r.text
-    db = SessionLocal()
-    try:
-        assert db.get(User, uid).force_onboarding is True
-        logs = db.execute(
-            select(AdminAuditLog).where(
-                AdminAuditLog.action == "user.force_onboarding.set",
-                AdminAuditLog.target_id == str(uid),
-            )
-        ).scalars().all()
-        assert logs[0].detail == {"before": False, "after": True}
-    finally:
-        db.close()
-    # 取消(回到 false)
-    assert admin_client.post(
-        f"/admin/api/users/{uid}/force-onboarding", json={"enabled": False},
-        headers=_auth(operator_token),
-    ).status_code == 200
-    db = SessionLocal()
-    try:
-        assert db.get(User, uid).force_onboarding is False
-    finally:
-        db.close()
-
-
-def test_finance_cannot_force_onboarding(admin_client: TestClient, finance_token: str) -> None:
-    uid = _seed_user("13900000010")
-    r = admin_client.post(
-        f"/admin/api/users/{uid}/force-onboarding", json={"enabled": True},
-        headers=_auth(finance_token),
-    )
-    assert r.status_code == 403
-
-
 # ===== 角色守卫 =====
 
 def test_operator_cannot_grant_coins(admin_client: TestClient, operator_token: str) -> None:

tests/test_onboarding.py

@@ -65,33 +65,3 @@ def test_onboarding_missing_device_id_treated_as_incomplete(client) -> None:
     phone = "13800138003"
     data = _login(client, phone, device_id=None)
     assert data["onboarding_completed"] is False
-
-
-def test_complete_clears_force_onboarding(client) -> None:
-    """运营「一键开启新手引导」(force_onboarding=True)后,用户走完引导即自动清除该标记,
-    /me 随之回到 false——否则下次启动会被无限拉回引导。"""
-    from app.db.session import SessionLocal
-    from app.models.user import User
-
-    phone = "13800138004"
-    dev = "android-id-force"
-    data = _login(client, phone, dev)
-    access = data["access_token"]
-    uid = data["user"]["id"]
-
-    # 模拟运营在后台一键开启(admin 接口本身另在 test_admin_write 覆盖)
-    db = SessionLocal()
-    try:
-        db.get(User, uid).force_onboarding = True
-        db.commit()
-    finally:
-        db.close()
-
-    def _me() -> dict:
-        return client.get(
-            "/api/v1/auth/me", headers={"Authorization": f"Bearer {access}"}
-        ).json()
-
-    assert _me()["force_onboarding"] is True
-    assert _mark_complete(client, access, dev).status_code == 200
-    assert _me()["force_onboarding"] is False