fix: account open fire-and-forget; open-browser direct+sync

SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: 账号与凭据管理
 description: "通用账号 & 凭据 & 登录流程管理器。管理网页/RPA 账号、API Key/Token；支持 9 种登录策略（账密自动、扫码、2FA、SSO 等）；提供加密本地存储（Fernet）；提供 rpa_helpers 库供业务 skill 同进程调用 ensure_logged_in 完成多平台自动登录。涵盖物流、LLM、内容、开发、API 五类平台。"
-version: 2.0.12
+version: 2.0.16
 author: 深圳匠厂科技有限公司
 metadata:
   openclaw:
@@ -27,19 +27,21 @@ allowed-tools:
 
 ## 何时使用本技能
 
-1. **登记网页/RPA 账号**：`account add-web` — 平台 + 登录标识 + 自动生成 profile_dir；可选 `--auth-strategy` / `--session-persistent` / `--device-fingerprint`。
-2. **登记凭据（API Key / Token / 加密密码）**：`account add-secret` — `none` / `env` / `windows_credential` / `local_encrypted`。
-3. **查看/列出账号**：`account get/list` — JSON 输出；list 可按 `--login-id` / `--label` 筛选。
-4. **用人话定位账号**：`account resolve --login-id <手机号/用户名>` 或 `--label` / `--query`（不必先查数字 id）。
-5. **为 RPA 挑选账号**：`account pick-web` — 按平台/环境/角色，也可 `--login-id` / `--label`；业务首启推荐 `account ensure-web`（或 `pick-web --ensure`）：无账号则自动 `add-web` 再返回。
-6. **获取 API Key（凭据表）**：`credential pick` — 默认隐藏 secret，`--reveal` 时才返回。
-7. **浏览器打开查看**：`account open <id>`（或兼容别名 `open <id>`）— 仅查看，不做站点侧登录判定。
-8. **管理 lease**：`lease release/list/cleanup` — 防止 RPA 并发。
-9. **初始化加密**：`account init-master-key` — 生成 Fernet master.key。
-10. **加密备份与迁移**：`account export-credentials --plaintext` / `account import-credentials` — 导出明文 JSON、在新环境重新加密写入。
-11. **按 lease 取明文密码**：`account get-credential <id> --reveal --lease-token <token>` — 结合 pick-web 返回的 lease 解密 `local_encrypted` 等路径。
-12. **业务 skill 同进程登录**：`from service.rpa_helpers import ensure_logged_in` — 详见 [`references/RPA_HELPERS.md`](references/RPA_HELPERS.md)。
-13. **有则用、无则建（网页账号）**：`account ensure-web --platform <p>` — 登记账号 ≠ 已登录；登录仍由业务技能 / `ensure_logged_in` 处理。
+1. **安装/升级后初始化库与数据管理元数据**：`init-db`（宿主会自动调用；也可手工执行）
+2. **登记网页/RPA 账号**：`account add-web` — 平台 + 登录标识 + 自动生成 profile_dir；可选 `--auth-strategy` / `--session-persistent` / `--device-fingerprint`。
+3. **登记凭据（API Key / Token / 加密密码）**：`account add-secret` — `none` / `env` / `windows_credential` / `local_encrypted`。
+4. **查看/列出账号**：`account get/list` — JSON 输出；list 可按 `--login-id` / `--label` 筛选。
+5. **用人话定位账号**：`account resolve --login-id <手机号/用户名>` 或 `--label` / `--query`（不必先查数字 id）。
+6. **为 RPA 挑选账号**：`account pick-web` — 按平台/环境/角色，也可 `--login-id` / `--label`；业务首启推荐 `account ensure-web`（或 `pick-web --ensure`）：无账号则自动 `add-web` 再返回。
+7. **获取 API Key（凭据表）**：`credential pick` — 默认隐藏 secret，`--reveal` 时才返回。
+8. **浏览器打开查看**：`account open <id>`（或兼容别名 `open <id>`）— 拉起 Profile 后立即返回，不做站点侧登录判定。数据管理「账号信息」行内 **打开浏览器** / **打开文件位置** 均为 `interaction: direct` + `sync`（见 `assets/actions.json`）。
+9. **打开 Profile 目录**：`account reveal-profile <id>` — 在系统文件管理器中打开该账号 `profile_dir`。
+10. **管理 lease**：`lease release/list/cleanup` — 防止 RPA 并发。
+11. **初始化加密**：`account init-master-key` — 生成 Fernet master.key。
+12. **加密备份与迁移**：`account export-credentials --plaintext` / `account import-credentials` — 导出明文 JSON、在新环境重新加密写入。
+13. **按 lease 取明文密码**：`account get-credential <id> --reveal --lease-token <token>` — 结合 pick-web 返回的 lease 解密 `local_encrypted` 等路径。
+14. **业务 skill 同进程登录**：`from service.rpa_helpers import ensure_logged_in` — 详见 [`references/RPA_HELPERS.md`](references/RPA_HELPERS.md)。
+15. **有则用、无则建（网页账号）**：`account ensure-web --platform <p>` — 登记账号 ≠ 已登录；登录仍由业务技能 / `ensure_logged_in` 处理。
 
 ## CLI 调用
 

assets/README.md

@@ -1,5 +1,6 @@
 # assets
 
+- `actions.json`：匠厂宿主 Skill Action 清单（数据管理「账号信息」行内：打开文件位置 / 打开浏览器）。
 - `examples/`：CLI 成功输出形状示例（虚构路径与数据）。
 - `schemas/`：`get` / `list-json` / `pick-web` 单条对象的轻量 JSON Schema（`additionalProperties: true` 允许扩展字段）。
 

assets/actions.json

@@ -0,0 +1,64 @@
+{
+  "schemaVersion": 1,
+  "skill": "account-manager",
+  "actions": [
+    {
+      "id": "reveal-profile",
+      "label": "打开文件位置",
+      "description": "在系统文件管理器中打开该账号的浏览器 Profile 目录。",
+      "placements": ["row"],
+      "executionProfile": "sync",
+      "interaction": "direct",
+      "bind": {
+        "tables": ["accounts"],
+        "inputMapping": { "accountId": "$row.$pk" }
+      },
+      "inputSchema": {
+        "type": "object",
+        "required": ["accountId"],
+        "properties": {
+          "accountId": {
+            "type": "number",
+            "title": "账号 ID",
+            "description": "accounts 表主键",
+            "readOnly": true
+          }
+        }
+      },
+      "entrypoint": {
+        "type": "cli",
+        "command": "account",
+        "args": ["reveal-profile", "{{accountId}}"]
+      }
+    },
+    {
+      "id": "open-browser",
+      "label": "打开浏览器",
+      "description": "用该账号持久化 Profile 打开 Chrome/Edge，仅供肉眼核对（不做登录判定）。一点即开，不等待关闭。",
+      "placements": ["row"],
+      "executionProfile": "sync",
+      "interaction": "direct",
+      "bind": {
+        "tables": ["accounts"],
+        "inputMapping": { "accountId": "$row.$pk" }
+      },
+      "inputSchema": {
+        "type": "object",
+        "required": ["accountId"],
+        "properties": {
+          "accountId": {
+            "type": "number",
+            "title": "账号 ID",
+            "description": "accounts 表主键",
+            "readOnly": true
+          }
+        }
+      },
+      "entrypoint": {
+        "type": "cli",
+        "command": "account",
+        "args": ["open", "{{accountId}}"]
+      }
+    }
+  ]
+}

references/CLI.md

@@ -17,6 +17,12 @@
 
 ## 子命令速查（按业务流程）
 
+### 0. 运行时初始化（宿主标准）
+
+| CLI | 说明 |
+|---|---|
+| `init-db` | 幂等初始化本地 SQLite、schema 迁移与数据管理中文元数据（`_jiangchang_*`）。成功输出 JSON：`{"ok":true,"skill":"account-manager","db_path":"..."}`。**匠厂安装/升级技能时会调用本命令**；业务命令内部也会触发同等 `init_db()`。 |
+
 ### A. 平台管理
 
 | CLI | 说明 |
@@ -48,7 +54,8 @@
 | `account mark-session <id> --session-status <status>` | 标记会话状态 |
 | `account mark-error <id> --code <code> --message <msg>` | 标记错误 |
 | `account mark-used <id>` | 标记已使用 |
-| `account open <id>` | 打开持久化浏览器 profile（仅查看） |
+| `account open <id>` | 打开持久化浏览器 profile（仅查看；**拉起后立即返回**，不等待关闭） |
+| `account reveal-profile <id>` | 在系统文件管理器中打开该账号 `profile_dir` |
 
 ### D. 加密备份与迁移（v2）
 
@@ -69,6 +76,7 @@
 | `lease list [--active]` | 列出租约 |
 | `lease cleanup` | 清理过期租约 |
 | `account open <id>` / `open <id>` | 浏览器查看 |
+| `account reveal-profile <id>` | 打开 Profile 目录 |
 
 ### 兼容入口（旧版别名）
 

references/ERRORS.md

@@ -87,6 +87,7 @@
 |---|---|
 | `ERROR:需要 playwright Python 包` | 未安装 `playwright` 包或本机无 Chrome/Edge |
 | `ERROR:PROFILE_DIR_MISSING` | 账号记录中 `profile_dir` 为空 |
+| `ERROR:REVEAL_PROFILE_FAILED` | `account reveal-profile` 无法打开目录（系统调用失败） |
 
 ---
 

scripts/cli/app.py

@@ -4,7 +4,7 @@ CLI entry point: argv dispatch for account-manager v2.
 
 Subcommand tree:
   platform  list/get/ensure
-  account   add-web|ensure-web|add-secret|get|get-credential|list|resolve|pick-web|mark-session|mark-error|mark-used|open
+  account   add-web|ensure-web|add-secret|get|get-credential|list|resolve|pick-web|mark-session|mark-error|mark-used|open|reveal-profile
             init-master-key|export-credentials|import-credentials
   credential pick|resolve
   lease     release|list|cleanup
@@ -38,6 +38,7 @@ from service.account_service import (
     cmd_delete_by_platform,
     cmd_delete_by_platform_phone,
     cmd_get,
+    cmd_init_db,
     cmd_lease_cleanup,
     cmd_lease_list,
     cmd_lease_release,
@@ -48,7 +49,7 @@ from service.account_service import (
     cmd_platform_get,
     cmd_platform_list,
 )
-from service.browser_service import cmd_open
+from service.browser_service import cmd_open, cmd_reveal_profile
 from util.constants import LOG_LOGGER_NAME, SKILL_SLUG
 from util.logging_config import get_skill_logger, setup_skill_logging
 from util.platforms import _platform_list_cn_for_help, resolve_platform_key
@@ -60,6 +61,9 @@ from util.runtime_paths import _maybe_print_paths_debug_cli
 
 _USAGE = """account-manager v2 — Credential & Browser Profile Manager
 
+运行时 / 数据管理:
+  init-db   # 幂等初始化本地库与数据管理中文元数据（宿主安装/升级会调用）
+
 平台管理:
   platform list [--domain logistics|llm|content]
   platform get <platform>
@@ -100,6 +104,7 @@ _USAGE = """account-manager v2 — Credential & Browser Profile Manager
   account mark-error <id> --code <code> --message <msg>
   account mark-used <id>
   account open <id>
+  account reveal-profile <id>
   account init-master-key [--print|--from-env|--rotate]
   account export-credentials --plaintext
   account import-credentials [--replace-all] < backup.json
@@ -114,8 +119,9 @@ _USAGE = """account-manager v2 — Credential & Browser Profile Manager
   lease list [--active]
   lease cleanup
 
-浏览器:
+浏览器 / Profile:
   account open <id>
+  account reveal-profile <id>   # 在文件管理器中打开 profile_dir
   open <id>   # 兼容别名
 
 兼容入口（旧版）:
@@ -217,6 +223,10 @@ def main() -> None:
     # Determine if first arg is a platform key (for legacy shorthand)
     plat_from_first = resolve_platform_key(cmd)
 
+    # ---- init-db（宿主 install/upgrade 标准入口）----
+    if cmd == "init-db":
+        sys.exit(cmd_init_db())
+
     # ---- platform subcommand ----
     if cmd == "platform":
         if len(av) < 3:
@@ -314,6 +324,11 @@ def main() -> None:
                 print("ERROR:CLI_BAD_ARGS account open 需要 <id>。")
                 sys.exit(1)
             cmd_open(av[3])
+        elif sub == "reveal-profile":
+            if len(av) < 4:
+                print("ERROR:CLI_BAD_ARGS account reveal-profile 需要 <id>。")
+                sys.exit(1)
+            cmd_reveal_profile(av[3])
         elif sub == "init-master-key":
             cmd_account_init_master_key(av)
         elif sub == "export-credentials":

scripts/service/account_service.py

@@ -39,7 +39,12 @@ from service.legacy_commands import (
     cmd_list,
 )
 from service.lease_commands import cmd_lease_cleanup, cmd_lease_list, cmd_lease_release
-from service.platform_commands import cmd_platform_ensure, cmd_platform_get, cmd_platform_list
+from service.platform_commands import (
+    cmd_init_db,
+    cmd_platform_ensure,
+    cmd_platform_get,
+    cmd_platform_list,
+)
 
 __all__ = [
     "PlatformCallerMeta",
@@ -60,6 +65,7 @@ __all__ = [
     "cmd_delete_by_platform",
     "cmd_delete_by_platform_phone",
     "cmd_get",
+    "cmd_init_db",
     "cmd_lease_cleanup",
     "cmd_lease_list",
     "cmd_lease_release",

scripts/service/browser_service.py

@@ -114,10 +114,85 @@ def resolve_account_browser_url(account: dict, conn=None) -> str:
     return (PLATFORMS.get(pk, {}).get("default_url") or "").strip()
 
 
+def cmd_reveal_profile(account_id):
+    """
+    在系统文件管理器中打开该账号的 profile_dir。
+    不写数据库、不启动浏览器。
+    """
+    log = get_skill_logger()
+    init_db()
+    log.info("reveal_profile_start account_id=%r", account_id)
+
+    target = get_account_by_id(account_id)
+    if not target:
+        log.warning("reveal_profile_error account_id=%r not_found", account_id)
+        print("ERROR:ACCOUNT_NOT_FOUND")
+        sys.exit(1)
+
+    profile_dir = (target.get("profile_dir") or "").strip()
+    if not profile_dir:
+        log.warning("reveal_profile_error account_id=%s profile_dir_missing", account_id)
+        print("ERROR:PROFILE_DIR_MISSING 账号中 profile_dir 为空。")
+        sys.exit(1)
+
+    profile_abs = os.path.abspath(profile_dir)
+    os.makedirs(profile_abs, exist_ok=True)
+
+    try:
+        if sys.platform == "win32":
+            # explorer 即使成功也常返回非 0；用 startfile 打开目录更稳妥
+            os.startfile(profile_abs)  # type: ignore[attr-defined]
+        elif sys.platform == "darwin":
+            subprocess.run(["open", profile_abs], check=False)
+        else:
+            subprocess.run(["xdg-open", profile_abs], check=False)
+    except OSError as exc:
+        log.error("reveal_profile_error account_id=%s error=%s", account_id, str(exc))
+        print(f"ERROR:REVEAL_PROFILE_FAILED {exc}")
+        sys.exit(1)
+
+    log.info("reveal_profile_ok account_id=%s profile_dir=%s", account_id, profile_abs)
+    print(
+        json.dumps(
+            {
+                "success": True,
+                "data": {
+                    "account_id": int(target["id"]) if target.get("id") is not None else account_id,
+                    "profile_dir": profile_abs,
+                },
+            },
+            ensure_ascii=False,
+        )
+    )
+
+
+def _spawn_browser_child(cmd: list[str], *, cwd: str, env: dict) -> subprocess.Popen:
+    """Detach Playwright child so `account open` can return immediately."""
+    popen_kwargs: dict = {
+        "cwd": cwd,
+        "env": env,
+        "stdin": subprocess.DEVNULL,
+        "stdout": subprocess.DEVNULL,
+        "stderr": subprocess.DEVNULL,
+        "close_fds": True,
+    }
+    if sys.platform == "win32":
+        # Keep browser window; detach from parent console / Skill Action wait.
+        flags = subprocess.CREATE_NEW_PROCESS_GROUP | subprocess.DETACHED_PROCESS
+        create_no_window = getattr(subprocess, "CREATE_NO_WINDOW", 0)
+        if create_no_window:
+            flags |= create_no_window
+        popen_kwargs["creationflags"] = flags
+    else:
+        popen_kwargs["start_new_session"] = True
+    return subprocess.Popen(cmd, **popen_kwargs)
+
+
 def cmd_open(account_id):
     """
     打开该账号的持久化浏览器，仅用于肉眼核对。
     不写数据库、不判定是否已登录。
+    拉起浏览器子进程后立即返回（不等待窗口关闭），便于 Skill Action sync/direct。
     """
     log = get_skill_logger()
     init_db()
@@ -127,26 +202,24 @@ def cmd_open(account_id):
     if not target:
         log.warning("browser_open_error account_id=%r not_found", account_id)
         print("ERROR:ACCOUNT_NOT_FOUND")
-        return
+        sys.exit(1)
 
     channel = resolve_chromium_channel()
     if not channel:
         _print_browser_install_hint()
-        return
+        sys.exit(1)
 
     try:
         import playwright  # noqa: F401
     except ImportError:
-        print(
-            "ERROR:需要 playwright Python 包。"
-        )
-        return
+        print("ERROR:需要 playwright Python 包。")
+        sys.exit(1)
 
     profile_dir = (target.get("profile_dir") or "").strip()
     if not profile_dir:
         log.warning("browser_open_error account_id=%s profile_dir_missing", account_id)
         print("ERROR:PROFILE_DIR_MISSING 账号中 profile_dir 为空。")
-        return
+        sys.exit(1)
     os.makedirs(profile_dir, exist_ok=True)
 
     conn = get_conn()
@@ -164,28 +237,49 @@ def cmd_open(account_id):
     )
 
     browser_name = "Google Chrome" if channel == "chrome" else "Microsoft Edge"
-    print(f"正在打开 [{target['account_label']}] 的 {browser_name}（仅查看，不写入数据库）")
-    print(f"地址：{url}")
-    print("本命令不做登录检测。")
-
     cfg = {"channel": channel, "profile_dir": profile_dir, "url": url}
     with tempfile.NamedTemporaryFile(
         mode="w", suffix=".json", delete=False, encoding="utf-8"
     ) as jf:
         json.dump(cfg, jf, ensure_ascii=False)
         cfg_path = jf.name
+
     try:
-        subprocess.run(
+        child = _spawn_browser_child(
             [sys.executable, "-m", "service.open_child_runner", cfg_path],
             cwd=_SCRIPTS_ROOT,
             env=subprocess_env_with_trace(),
         )
-        log.info("browser_open_close account_id=%s channel=%s", account_id, channel)
     except Exception as e:
-        log.error("browser_open_error account_id=%s error=%s", account_id, str(e))
-        raise
-    finally:
         try:
             os.unlink(cfg_path)
         except OSError:
             pass
+        log.error("browser_open_error account_id=%s error=%s", account_id, str(e))
+        print(f"ERROR:BROWSER_OPEN_FAILED {e}")
+        sys.exit(1)
+
+    log.info(
+        "browser_open_spawned account_id=%s channel=%s pid=%s",
+        account_id,
+        channel,
+        child.pid,
+    )
+    print(
+        json.dumps(
+            {
+                "success": True,
+                "data": {
+                    "account_id": int(target["id"]) if target.get("id") is not None else account_id,
+                    "account_label": target.get("account_label"),
+                    "browser": browser_name,
+                    "channel": channel,
+                    "profile_dir": os.path.abspath(profile_dir),
+                    "url": url,
+                    "detached": True,
+                    "pid": child.pid,
+                },
+            },
+            ensure_ascii=False,
+        )
+    )

scripts/service/open_child_runner.py

@@ -1,5 +1,9 @@
-"""Playwright 仅打开浏览器子进程：由 browser_service.cmd_open 启动，argv[1] 为 JSON 配置路径。"""
+"""Playwright 仅打开浏览器子进程：由 browser_service.cmd_open 启动，argv[1] 为 JSON 配置路径。
+
+父进程拉起后即返回；本子进程自行维护浏览器生命周期，读完配置后删除临时 JSON。
+"""
 import json
+import os
 import sys
 
 from playwright.sync_api import sync_playwright
@@ -12,8 +16,14 @@ from util.playwright_stealth import (
 
 
 def main():
-    with open(sys.argv[1], encoding="utf-8") as f:
+    cfg_path = sys.argv[1]
+    with open(cfg_path, encoding="utf-8") as f:
         c = json.load(f)
+    try:
+        os.unlink(cfg_path)
+    except OSError:
+        pass
+
     with sync_playwright() as p:
         args, ignore_automation = persistent_context_launch_parts()
         lc_kw = dict(

scripts/service/platform_commands.py

@@ -6,8 +6,38 @@ from typing import Optional
 from db.accounts_repo import get_platform_from_db, list_platforms_from_db, upsert_platform
 from db.connection import get_conn, init_db
 from service._svc_common import _err_json, _ok_json, _resolve_or_fail, _say
+from util.constants import SKILL_SLUG
 from util.logging_config import get_skill_logger
 from util.platforms import DEFAULT_CAPABILITIES_JSON, is_valid_platform_key_format, seed_platforms
+from util.runtime_paths import get_db_path
+
+
+def cmd_init_db() -> int:
+    """幂等创建/迁移本地 SQLite 与数据管理元数据（宿主 install/upgrade 可调用）。"""
+    log = get_skill_logger()
+    log.info("cli_start subcommand=init_db")
+    try:
+        init_db()
+    except Exception as exc:  # noqa: BLE001 — CLI 边界，统一成可机读失败
+        log.exception("init_db_failure")
+        print(
+            json.dumps(
+                {"ok": False, "skill": SKILL_SLUG, "error": str(exc)},
+                ensure_ascii=False,
+            )
+        )
+        return 1
+    print(
+        json.dumps(
+            {
+                "ok": True,
+                "skill": SKILL_SLUG,
+                "db_path": get_db_path(),
+            },
+            ensure_ascii=False,
+        )
+    )
+    return 0
 
 
 def cmd_platform_list(domain: Optional[str] = None) -> None:

tests/test_browser_open.py

@@ -14,8 +14,7 @@ class TestPlaywrightLaunchHelpers:
         args, _ignore = persistent_context_launch_parts()
         assert "--start-maximized" in args
 
-    def test_open_child_runner_launch_keywords_documented(self):
-        """launch_persistent_context kwargs contract (static read of source)."""
+    def test_open_child_runner_cleans_config_after_read(self):
         import os
 
         path = os.path.join(
@@ -28,10 +27,63 @@ class TestPlaywrightLaunchHelpers:
         path = os.path.abspath(path)
         with open(path, encoding="utf-8") as f:
             src = f.read()
+        assert "os.unlink(cfg_path)" in src
+        assert "wait_for_event" in src
         assert "launch_persistent_context" in src
         assert "headless=False" in src
         assert "no_viewport=True" in src
-        assert '"channel"' in src or "'channel'" in src
+
+
+class TestCmdOpenDetached:
+    def test_cmd_open_spawns_and_returns(self, clean_db, tmp_path, capsys, monkeypatch):
+        from db.accounts_repo import insert_account
+        from service import browser_service
+        from util.platforms import seed_platforms
+
+        seed_platforms(clean_db)
+        profile = tmp_path / "prof"
+        aid = insert_account(
+            clean_db,
+            platform_key="sohu",
+            account_label="open-test",
+            login_id="1",
+            login_id_type="phone",
+            tenant_id=None,
+            environment="production",
+            role="default",
+            provider_code="sohu",
+            profile_dir=str(profile),
+            url="https://example.com",
+            extra_json=None,
+            now=1_700_000_000,
+            auth_strategy="password_auto",
+        )
+        clean_db.commit()
+
+        monkeypatch.setattr(browser_service, "resolve_chromium_channel", lambda: "chrome")
+
+        class FakeProc:
+            pid = 4242
+
+        spawned = {}
+
+        def fake_spawn(cmd, *, cwd, env):
+            spawned["cmd"] = cmd
+            spawned["cwd"] = cwd
+            return FakeProc()
+
+        monkeypatch.setattr(browser_service, "_spawn_browser_child", fake_spawn)
+
+        browser_service.cmd_open(str(aid))
+        out = capsys.readouterr().out.strip()
+        import json
+
+        payload = json.loads(out.splitlines()[-1])
+        assert payload["success"] is True
+        assert payload["data"]["detached"] is True
+        assert payload["data"]["pid"] == 4242
+        assert "-m" in spawned["cmd"]
+        assert "service.open_child_runner" in spawned["cmd"]
 
 
 class TestResolveChromiumChannel:

tests/test_reveal_profile.py

@@ -0,0 +1,75 @@
+# -*- coding: utf-8 -*-
+"""account reveal-profile：打开 profile 目录（不启动浏览器）。"""
+from __future__ import annotations
+
+import json
+import os
+from unittest import mock
+
+import pytest
+
+
+def _insert_account(conn, *, profile_dir: str) -> int:
+    from db.accounts_repo import insert_account
+    from util.platforms import seed_platforms
+
+    seed_platforms(conn)
+    aid = insert_account(
+        conn,
+        platform_key="sohu",
+        account_label="reveal-test",
+        login_id="13800000000",
+        login_id_type="phone",
+        tenant_id=None,
+        environment="production",
+        role="default",
+        provider_code="sohu",
+        profile_dir=profile_dir,
+        url="https://mp.sohu.com",
+        extra_json=None,
+        now=1_700_000_000,
+        auth_strategy="password_auto",
+        session_persistent=1,
+        device_fingerprint=None,
+    )
+    conn.commit()
+    return aid
+
+
+def test_reveal_profile_opens_dir(clean_db, tmp_path, capsys):
+    from service.browser_service import cmd_reveal_profile
+
+    profile = tmp_path / "profiles" / "sohu" / "reveal-test"
+    aid = _insert_account(clean_db, profile_dir=str(profile))
+
+    with mock.patch("service.browser_service.os.startfile") as startfile:
+        with mock.patch("service.browser_service.sys.platform", "win32"):
+            cmd_reveal_profile(str(aid))
+
+    startfile.assert_called_once()
+    called_path = startfile.call_args[0][0]
+    assert os.path.isdir(called_path)
+    assert os.path.abspath(str(profile)) == called_path
+
+    out = capsys.readouterr().out.strip()
+    payload = json.loads(out.splitlines()[-1])
+    assert payload["success"] is True
+    assert payload["data"]["account_id"] == aid
+    assert payload["data"]["profile_dir"] == called_path
+
+
+def test_reveal_profile_missing_account(clean_db):
+    from service.browser_service import cmd_reveal_profile
+
+    with pytest.raises(SystemExit) as ei:
+        cmd_reveal_profile("999999")
+    assert ei.value.code == 1
+
+
+def test_reveal_profile_empty_dir(clean_db):
+    from service.browser_service import cmd_reveal_profile
+
+    aid = _insert_account(clean_db, profile_dir="")
+    with pytest.raises(SystemExit) as ei:
+        cmd_reveal_profile(str(aid))
+    assert ei.value.code == 1

tests/test_usage.py

@@ -23,6 +23,7 @@ def test_usage_includes_all_account_subcommands():
     assert r.returncode == 1
     usage = r.stdout
     for cmd in (
+        "init-db",
         "init-master-key",
         "export-credentials",
         "import-credentials",
@@ -31,7 +32,49 @@ def test_usage_includes_all_account_subcommands():
         "get-credential",
         "pick-web",
         "resolve",
+        "reveal-profile",
     ):
         assert cmd in usage, f"USAGE 缺少 {cmd}"
     assert "--ensure" in usage
     assert "--login-id" in usage
+
+
+def test_init_db_cli_succeeds(fake_env):
+    """宿主调用的 init-db 必须成功并写入展示元数据。"""
+    import json
+
+    env = {
+        **os.environ,
+        "PYTHONPATH": _scripts,
+        "JIANGCHANG_DATA_ROOT": fake_env["data_root"],
+        "JIANGCHANG_USER_ID": fake_env["user_id"],
+    }
+    r = subprocess.run(
+        [sys.executable, "main.py", "init-db"],
+        capture_output=True,
+        text=True,
+        env=env,
+        cwd=_scripts,
+    )
+    assert r.returncode == 0, r.stdout + r.stderr
+    payload = json.loads(r.stdout.strip().splitlines()[-1])
+    assert payload.get("ok") is True
+    assert payload.get("skill") == "account-manager"
+    assert payload.get("db_path")
+
+    from db.connection import get_conn
+
+    conn = get_conn()
+    try:
+        cur = conn.cursor()
+        cur.execute("SELECT COUNT(*) FROM _jiangchang_columns")
+        assert int(cur.fetchone()[0]) > 0
+        cur.execute(
+            """
+            SELECT display_name FROM _jiangchang_columns
+            WHERE table_name = 'accounts' AND column_name = 'role'
+            """
+        )
+        assert cur.fetchone()[0] == "角色"
+    finally:
+        conn.close()