aria-code 4.1.3__py3-none-any.whl

setup_wizard.py

@@ -0,0 +1,653 @@
+#!/usr/bin/env python3
+"""
+setup_wizard.py — Aria Code 首次配置向导
+==========================================
+运行方式:
+  python3 setup_wizard.py             # 完整向导
+  python3 setup_wizard.py --model     # 仅配置模型
+  python3 setup_wizard.py --feishu    # 仅配置飞书
+  python3 setup_wizard.py --telegram  # 仅配置 Telegram
+
+向导完成后会生成:
+  ~/.aria/.env          环境变量配置
+  ~/.aria/config.json   Aria Code 配置
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+import os
+import platform
+import subprocess
+import sys
+import time
+import uuid
+from pathlib import Path
+
+# ── 尝试用 rich 渲染，回落到纯文本 ──────────────────────────────────────────
+
+try:
+    from rich.console import Console
+    from rich.panel import Panel
+    from rich.prompt import Confirm, Prompt
+    from rich.table import Table
+    from rich import box as rbox
+    _rich = True
+    console = Console()
+except ImportError:
+    _rich = False
+    class _FakeConsole:
+        def print(self, *a, **k): print(*a)
+        def rule(self, *a, **k): print("─" * 60)
+    console = _FakeConsole()  # type: ignore
+
+_ARIA_DIR = Path.home() / ".aria"
+_ENV_FILE = _ARIA_DIR / ".env"
+_CFG_FILE = _ARIA_DIR / "config.json"
+
+# ── System language detection ─────────────────────────────────────────────────
+
+try:
+    from apps.cli.i18n import detect_system_lang as _detect_lang
+    _SYS_LANG = _detect_lang()
+except Exception:
+    _SYS_LANG = "en"
+
+# Bilingual wizard strings — keyed by (_SYS_LANG, key)
+_WZ: dict[str, dict[str, str]] = {
+    "setup_model_title":     {"zh": "本地模型配置", "en": "Local Model Setup"},
+    "ollama_not_found":      {"zh": "未检测到 Ollama", "en": "Ollama not found"},
+    "install_ollama_q":      {"zh": "是否自动安装 Ollama？", "en": "Auto-install Ollama?"},
+    "install_ollama_ing":    {"zh": "正在安装 Ollama...", "en": "Installing Ollama..."},
+    "install_ollama_fail":   {"zh": "Ollama 安装失败，请手动安装: https://ollama.com",
+                              "en": "Ollama install failed. Visit https://ollama.com"},
+    "install_ollama_ok":     {"zh": "Ollama 安装完成", "en": "Ollama installed"},
+    "skip_ollama":           {"zh": "跳过 Ollama 安装。你也可以配置 API 模型（下方步骤）。",
+                              "en": "Skipping Ollama. You can configure a cloud API model below."},
+    "win_ollama_hint":       {"zh": "请访问 https://ollama.com/download 下载 Windows 版本后重新运行向导。",
+                              "en": "Download the Windows installer from https://ollama.com/download and re-run."},
+    "download_model":        {"zh": "正在下载模型 {name}，可能需要几分钟...",
+                              "en": "Downloading {name}, this may take a few minutes..."},
+    "select_model_prompt":   {"zh": "选择模型（输入序号/名称，直接回车使用 {default}）",
+                              "en": "Select model (number/name, Enter to use {default})"},
+    "select_model_no_local": {"zh": "选择模型（输入序号 1-6 或完整名称）",
+                              "en": "Select model (enter number 1-6 or full name)"},
+    "model_set":             {"zh": "已选择模型: {name}", "en": "Model set: {name}"},
+    "pull_model_q":          {"zh": "是否立即下载？", "en": "Download now?"},
+    "pull_skip":             {"zh": "跳过下载，请稍后运行: ollama pull {name}",
+                              "en": "Skipping download. Run later: ollama pull {name}"},
+    "lang_detected":         {"zh": "检测到系统语言：中文（可用 /config set ui_lang=en 切换）",
+                              "en": "Detected system language: English (use /config set ui_lang=zh to switch)"},
+    "api_keys_section":      {"zh": "API 密钥配置（可选）", "en": "API Keys (Optional)"},
+    "api_keys_info":         {"zh": "配置后可使用云端模型（Claude / GPT-4 / DeepSeek 等）",
+                              "en": "Configure to use cloud models (Claude / GPT-4 / DeepSeek etc.)"},
+    "api_key_prompt":        {"zh": "配置 {name} API Key{suffix}？",
+                              "en": "Configure {name} API Key{suffix}?"},
+    "api_key_saved":         {"zh": "{name} API Key 已保存", "en": "{name} API Key saved"},
+    "api_keys_done":         {"zh": "共配置 {count} 个 API Key",
+                              "en": "{count} API key(s) configured"},
+}
+
+
+def _wz(key: str, **kwargs) -> str:
+    """Look up wizard string in detected system language."""
+    entry = _WZ.get(key, {})
+    tmpl = entry.get(_SYS_LANG) or entry.get("en") or key
+    return tmpl.format(**kwargs) if kwargs else tmpl
+
+
+# ── Helpers ──────────────────────────────────────────────────────────────────
+
+def _ask(prompt: str, default: str = "", password: bool = False) -> str:
+    if _rich:
+        return Prompt.ask(prompt, default=default, password=password)
+    d = f" [{default}]" if default else ""
+    val = input(f"{prompt}{d}: ").strip()
+    return val or default
+
+
+def _confirm(prompt: str, default: bool = True) -> bool:
+    if _rich:
+        return Confirm.ask(prompt, default=default)
+    ans = input(f"{prompt} [{'Y/n' if default else 'y/N'}]: ").strip().lower()
+    return (ans != "n") if default else (ans == "y")
+
+
+def _section(title: str) -> None:
+    if _rich:
+        console.rule(f"[bold cyan]{title}[/bold cyan]")
+    else:
+        print(f"\n{'─'*20} {title} {'─'*20}")
+
+
+def _ok(msg: str)   -> None: console.print(f"[green]✓[/green] {msg}" if _rich else f"✓ {msg}")
+def _warn(msg: str) -> None: console.print(f"[yellow]⚠[/yellow] {msg}" if _rich else f"⚠ {msg}")
+def _err(msg: str)  -> None: console.print(f"[red]✗[/red] {msg}" if _rich else f"✗ {msg}")
+def _info(msg: str) -> None: console.print(f"[dim]{msg}[/dim]" if _rich else msg)
+
+
+def _load_env() -> dict[str, str]:
+    env: dict[str, str] = {}
+    if _ENV_FILE.exists():
+        for line in _ENV_FILE.read_text().splitlines():
+            line = line.strip()
+            if line and not line.startswith("#") and "=" in line:
+                k, _, v = line.partition("=")
+                env[k.strip()] = v.strip()
+    return env
+
+
+def _save_env(env: dict[str, str]) -> None:
+    _ARIA_DIR.mkdir(parents=True, exist_ok=True)
+    lines = ["# Aria Code 配置 — 由 setup_wizard.py 生成\n"]
+    for k, v in sorted(env.items()):
+        lines.append(f"{k}={v}\n")
+    _ENV_FILE.write_text("".join(lines))
+    _ENV_FILE.chmod(0o600)
+
+
+# ── Step 1: Ollama check + model selection ───────────────────────────────────
+
+def _ollama_installed() -> bool:
+    return subprocess.run(["which", "ollama"], capture_output=True).returncode == 0
+
+
+def _ollama_models() -> list[str]:
+    try:
+        result = subprocess.run(
+            ["ollama", "list"], capture_output=True, text=True, timeout=10
+        )
+        lines = result.stdout.strip().splitlines()[1:]  # skip header
+        return [l.split()[0] for l in lines if l.strip()]
+    except Exception:
+        return []
+
+
+def _install_ollama() -> bool:
+    _info(_wz("install_ollama_ing"))
+    system = platform.system()
+    if system == "Darwin":
+        if subprocess.run(["which", "brew"], capture_output=True).returncode == 0:
+            ret = subprocess.run(["brew", "install", "ollama"]).returncode
+            return ret == 0
+        cmd = 'curl -fsSL https://ollama.com/install.sh | sh'
+    elif system == "Linux":
+        cmd = 'curl -fsSL https://ollama.com/install.sh | sh'
+    elif system == "Windows":
+        _warn(_wz("win_ollama_hint"))
+        return False
+    else:
+        _warn(f"Unknown system {system}. Install manually: https://ollama.com")
+        return False
+    return subprocess.run(cmd, shell=True).returncode == 0
+
+
+def _pull_model(name: str) -> bool:
+    _info(_wz("download_model", name=name))
+    return subprocess.run(["ollama", "pull", name]).returncode == 0
+
+
+_RECOMMENDED_MODELS_ZH = [
+    ("qwen2.5:7b",           "阿里通义千问 7B  ·  中文最强 · ~4GB  · 推荐"),
+    ("deepseek-r1:7b",       "DeepSeek R1 7B   ·  推理强   · ~4GB"),
+    ("llama3.2:3b",          "Meta Llama 3.2 3B ·  速度快   · ~2GB"),
+    ("mistral:7b",           "Mistral 7B        ·  均衡     · ~4GB"),
+    ("qwen2.5:14b",          "通义千问 14B      ·  质量高   · ~8GB  · 需 16GB RAM"),
+    ("deepseek-r1:14b",      "DeepSeek R1 14B   ·  推理最强 · ~8GB  · 需 16GB RAM"),
+]
+_RECOMMENDED_MODELS_EN = [
+    ("qwen2.5:7b",           "Qwen2.5 7B  ·  Best Chinese+English · ~4GB  · Recommended"),
+    ("deepseek-r1:7b",       "DeepSeek R1 7B   ·  Strong reasoning · ~4GB"),
+    ("llama3.2:3b",          "Meta Llama 3.2 3B ·  Fast, lightweight · ~2GB"),
+    ("mistral:7b",           "Mistral 7B        ·  Balanced · ~4GB"),
+    ("qwen2.5:14b",          "Qwen2.5 14B       ·  High quality · ~8GB  · Needs 16GB RAM"),
+    ("deepseek-r1:14b",      "DeepSeek R1 14B   ·  Best reasoning · ~8GB · Needs 16GB RAM"),
+]
+_RECOMMENDED_MODELS = _RECOMMENDED_MODELS_ZH if _SYS_LANG == "zh" else _RECOMMENDED_MODELS_EN
+
+
+def setup_model(env: dict[str, str]) -> None:
+    _section(_wz("setup_model_title"))
+
+    if not _ollama_installed():
+        _warn(_wz("ollama_not_found"))
+        if _confirm(_wz("install_ollama_q")):
+            if not _install_ollama():
+                _err(_wz("install_ollama_fail"))
+                return
+            _ok(_wz("install_ollama_ok"))
+        else:
+            _info(_wz("skip_ollama"))
+            return
+
+    existing = _ollama_models()
+    _installed_label = "✓ installed" if _SYS_LANG != "zh" else "✓ 已安装"
+    _other_label = "Other installed model" if _SYS_LANG != "zh" else "已安装的其他模型"
+
+    if _rich:
+        _col_model  = "Model" if _SYS_LANG != "zh" else "模型"
+        _col_desc   = "Description" if _SYS_LANG != "zh" else "说明"
+        t = Table(box=rbox.SIMPLE, show_header=True)
+        t.add_column("#", style="dim", width=3)
+        t.add_column(_col_model, style="cyan")
+        t.add_column(_col_desc, style="dim")
+        for i, (name, desc) in enumerate(_RECOMMENDED_MODELS, 1):
+            marker = f" [green]{_installed_label}[/green]" if name in existing else ""
+            t.add_row(str(i), name + marker, desc)
+        if existing:
+            for m in existing:
+                if not any(m == n for n, _ in _RECOMMENDED_MODELS):
+                    t.add_row("*", m, _other_label)
+        console.print(t)
+    else:
+        for i, (name, desc) in enumerate(_RECOMMENDED_MODELS, 1):
+            installed = f" [{_installed_label}]" if name in existing else ""
+            print(f"  {i}. {name}{installed} — {desc}")
+
+    if existing:
+        default_model = existing[0]
+        choice = _ask(_wz("select_model_prompt", default=default_model), default=default_model)
+    else:
+        choice = _ask(_wz("select_model_no_local"), default="1")
+
+    # Resolve number → model name
+    try:
+        idx = int(choice) - 1
+        if 0 <= idx < len(_RECOMMENDED_MODELS):
+            model_name = _RECOMMENDED_MODELS[idx][0]
+        else:
+            model_name = choice
+    except ValueError:
+        model_name = choice
+
+    if model_name not in existing:
+        _pull_q = (f"模型 {model_name} 未安装，是否立即下载？" if _SYS_LANG == "zh"
+                   else f"Model {model_name} not installed. Download now?")
+        if _confirm(_pull_q):
+            if _pull_model(model_name):
+                _ok(f"模型 {model_name} 下载完成" if _SYS_LANG == "zh"
+                    else f"Model {model_name} downloaded")
+            else:
+                _err(_wz("pull_skip", name=model_name))
+        else:
+            _warn(_wz("pull_skip", name=model_name))
+
+    env["ARIA_DEFAULT_MODEL"] = model_name
+    env["OLLAMA_BASE_URL"] = "http://localhost:11434"
+    _ok(_wz("model_set", name=model_name))
+
+
+# ── Step 2: API keys (optional) ──────────────────────────────────────────────
+
+def setup_api_keys(env: dict[str, str]) -> None:
+    _section(_wz("api_keys_section"))
+    _info(_wz("api_keys_info"))
+
+    # (display_name, env_var, hint)
+    _LLM_PROVIDERS = [
+        # ── 国际
+        ("Anthropic (Claude)",         "ANTHROPIC_API_KEY",     "claude.ai/settings/keys"),
+        ("OpenAI (GPT-4o / Whisper)",  "OPENAI_API_KEY",        "platform.openai.com/api-keys"),
+        ("DeepSeek",                   "DEEPSEEK_API_KEY",       "platform.deepseek.com"),
+        ("Google Gemini",              "GOOGLE_API_KEY",         "aistudio.google.com/apikey"),
+        ("xAI Grok",                   "XAI_API_KEY",            "console.x.ai"),
+        ("Groq (fast inference)",      "GROQ_API_KEY",           "console.groq.com/keys"),
+        ("Mistral AI",                 "MISTRAL_API_KEY",        "console.mistral.ai/api-keys"),
+        ("Cohere",                     "COHERE_API_KEY",         "dashboard.cohere.com/api-keys"),
+        ("Perplexity",                 "PERPLEXITY_API_KEY",     "perplexity.ai/settings/api"),
+        ("Together AI",                "TOGETHER_API_KEY",       "api.together.xyz/settings/api-keys"),
+        # ── 国内
+        ("SiliconFlow 硅基流动",        "SILICONFLOW_API_KEY",   "cloud.siliconflow.cn"),
+        ("DashScope 阿里云百炼",         "DASHSCOPE_API_KEY",     "dashscope.aliyuncs.com"),
+        ("Moonshot Kimi 月之暗面",       "MOONSHOT_API_KEY",      "platform.moonshot.cn/api-keys"),
+        ("Zhipu GLM 智谱",              "ZHIPU_API_KEY",          "open.bigmodel.cn"),
+        ("Baidu ERNIE 百度千帆",         "QIANFAN_ACCESS_KEY",    "console.bce.baidu.com/iam/#/iam/accesslist"),
+        ("ByteDance Doubao 字节豆包",    "ARK_API_KEY",            "console.volcengine.com/ark"),
+        ("MiniMax",                    "MINIMAX_API_KEY",        "platform.minimaxi.com"),
+        ("StepFun 阶跃星辰",            "STEPFUN_API_KEY",        "platform.stepfun.com"),
+        ("01.AI Yi 零一万物",            "ONEAI_API_KEY",          "platform.lingyiwanwu.com"),
+    ]
+
+    _configured = []
+    for name, env_var, hint in _LLM_PROVIDERS:
+        already = bool(env.get(env_var) or __import__("os").getenv(env_var))
+        suffix = f" [已配置]" if already else ""
+        prompt = _wz("api_key_prompt", name=name, suffix=suffix)
+        if _confirm(prompt, default=False):
+            _info(f"  → {hint}")
+            key = _ask(env_var, password=True)
+            if key:
+                env[env_var] = key
+                _configured.append(name)
+                _ok(_wz("api_key_saved", name=name))
+
+    if _configured:
+        _ok(_wz("api_keys_done", count=len(_configured)))
+
+
+# ── Step 3: Feishu connection ─────────────────────────────────────────────────
+
+def setup_feishu(env: dict[str, str]) -> None:
+    _section("飞书连接配置")
+
+    if _rich:
+        console.print(Panel(
+            "[bold]两种连接方式：[/bold]\n\n"
+            "[cyan]1. Aria 中继服务[/cyan]  [dim]（推荐）[/dim]\n"
+            "   • 无需创建飞书应用\n"
+            "   • 在飞书向 Aria 官方 Bot 发送绑定码即可\n"
+            "   • 消息经中继服务器转发到你的本机\n\n"
+            "[cyan]2. 自建飞书应用[/cyan]  [dim]（完全自主）[/dim]\n"
+            "   • 需要飞书开发者账号\n"
+            "   • 消息直接推送到你的服务器\n"
+            "   • 适合有公网 IP / 服务器的用户",
+            title="飞书连接方式",
+            border_style="cyan",
+        ))
+    else:
+        print("  1. Aria 中继服务（推荐，无需创建飞书应用）")
+        print("  2. 自建飞书应用（需要飞书开发者账号）")
+
+    choice = _ask("选择方式", default="1")
+
+    if choice == "1":
+        _setup_feishu_relay(env)
+    else:
+        _setup_feishu_own_app(env)
+
+
+def _setup_feishu_relay(env: dict[str, str]) -> None:
+    """Connect via Aria Relay Server — no Feishu developer account needed."""
+    relay_url = env.get("ARIA_RELAY_URL", "wss://relay.aria.ai")
+    relay_url = _ask("中继服务器地址", default=relay_url)
+    env["ARIA_RELAY_URL"] = relay_url
+    env["ARIA_RELAY_MODE"] = "relay"
+
+    # Generate or reuse client_id
+    client_id = env.get("ARIA_RELAY_CLIENT_ID") or f"aria-{uuid.uuid4().hex[:12]}"
+    env["ARIA_RELAY_CLIENT_ID"] = client_id
+
+    if _rich:
+        console.print(Panel(
+            f"[bold]你的绑定码：[/bold]\n\n"
+            f"[bold yellow on black]  ARIA-BIND-{client_id.upper()}  [/bold yellow on black]\n\n"
+            f"[dim]操作步骤：[/dim]\n"
+            f"1. 在飞书搜索 [cyan]Aria Bot[/cyan] 并添加好友\n"
+            f"2. 发送：[cyan]/bind ARIA-BIND-{client_id.upper()}[/cyan]\n"
+            f"3. 收到 [green]\"绑定成功\"[/green] 后，飞书消息将转发到你的 Aria\n\n"
+            f"[dim]（中继客户端会在 daemon 启动时自动连接）[/dim]",
+            title="📱 飞书绑定步骤",
+            border_style="yellow",
+        ))
+    else:
+        print(f"\n你的绑定码: ARIA-BIND-{client_id.upper()}")
+        print("1. 在飞书搜索 Aria Bot 并添加好友")
+        print(f"2. 发送: /bind ARIA-BIND-{client_id.upper()}")
+        print("3. 收到绑定成功后继续\n")
+
+    if _confirm("是否已完成绑定？（可以稍后运行向导再完成）", default=False):
+        _ok("飞书中继绑定完成")
+    else:
+        _warn("请完成绑定后再启动 daemon，否则飞书消息无法到达本机")
+
+
+def _setup_feishu_own_app(env: dict[str, str]) -> None:
+    """Self-hosted Feishu app configuration with step-by-step guidance."""
+    if _rich:
+        console.print(
+            "\n[dim]飞书开发者后台：https://open.feishu.cn/app[/dim]\n"
+            "步骤：创建自建应用 → 凭证与基本信息 → 复制 App ID 和 App Secret\n"
+        )
+    else:
+        print("飞书开发者后台：https://open.feishu.cn/app")
+
+    app_id = _ask("FEISHU_APP_ID (如: cli_xxxxxxxxxxxx)", default=env.get("FEISHU_APP_ID", ""))
+    secret = _ask("FEISHU_APP_SECRET", default=env.get("FEISHU_APP_SECRET", ""), password=True)
+
+    if app_id:
+        env["FEISHU_APP_ID"] = app_id
+    if secret:
+        env["FEISHU_APP_SECRET"] = secret
+    env["ARIA_RELAY_MODE"] = "own_app"
+
+    # Webhook URL guidance
+    _info("\n事件订阅 Request URL: http://<你的IP或域名>/api/v1/feishu/event")
+    _info("如果没有公网 IP，可以用 cloudflared 建立隧道：")
+    _info("  brew install cloudflared && cloudflared tunnel --url http://localhost:8000")
+
+    if _confirm("是否配置 FEISHU_WEBHOOK_URL（群机器人单向推送）？", default=False):
+        wh = _ask("FEISHU_WEBHOOK_URL")
+        if wh:
+            env["FEISHU_WEBHOOK_URL"] = wh
+            _ok("飞书 Webhook 已保存")
+
+    _ok("飞书自建应用配置完成")
+
+
+# ── Step 4: Telegram Bot ─────────────────────────────────────────────────────
+
+def setup_telegram(env: dict[str, str]) -> None:
+    _section("Telegram Bot 配置")
+
+    if _rich:
+        console.print(Panel(
+            "[bold]三步完成 Telegram Bot 接入：[/bold]\n\n"
+            "[cyan]第一步：创建 Bot，获取 Token[/cyan]\n"
+            "  1. 打开 Telegram，搜索并进入 [bold]@BotFather[/bold]\n"
+            "  2. 发送命令：/newbot\n"
+            "  3. 按提示填写 Bot 名称（如 Aria Alerts）和用户名（需以 bot 结尾）\n"
+            "  4. 复制 BotFather 回复中的 [bold]Token[/bold]\n"
+            "     格式：[dim]1234567890:AAxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx[/dim]\n\n"
+            "[cyan]第二步：获取你的 Chat ID[/cyan]\n"
+            "  1. 搜索并进入 [bold]@userinfobot[/bold]\n"
+            "  2. 发送任意消息，复制返回的 [bold]Id[/bold] 数字\n"
+            "  3. 多个用户用英文逗号分隔，如：[dim]123456,789012[/dim]\n\n"
+            "[cyan]第三步：启动 Bot[/cyan]\n"
+            "  • 运行 [bold]python3 aria_daemon.py[/bold]\n"
+            "  • 在 Telegram 向你的 Bot 发送 /start\n"
+            "  • 支持命令：/price AAPL · /brief · /alerts · /help",
+            title="🤖 Telegram Bot 配置指南",
+            border_style="cyan",
+        ))
+    else:
+        print("\n=== Telegram Bot 配置 ===")
+        print("步骤：")
+        print("  1. Telegram → 搜索 @BotFather → /newbot → 复制 Token")
+        print("  2. 搜索 @userinfobot → 发消息 → 复制 Id 数字")
+        print("  3. 填入下方后运行: python3 aria_daemon.py")
+
+    # ── 获取 Token ────────────────────────────────────────────────────────────
+    existing_token = env.get("TELEGRAM_BOT_TOKEN", "")
+    if existing_token and existing_token != "your_bot_token_here":
+        _info(f"当前 Token: {existing_token[:10]}…（留空保留）")
+    token = _ask("TELEGRAM_BOT_TOKEN", default=existing_token)
+
+    if not token or token == "your_bot_token_here":
+        _warn("未填写 Token，Telegram Bot 不会启动")
+        return
+
+    # ── 验证 Token ────────────────────────────────────────────────────────────
+    _info("验证 Token 中…")
+    bot_info = _verify_telegram_token(token)
+    if bot_info:
+        _ok(f"Token 有效！Bot 名称: @{bot_info.get('username', '?')}")
+    else:
+        _warn("Token 验证失败（可能是网络问题），继续保存但请检查 Token 是否正确")
+
+    env["TELEGRAM_BOT_TOKEN"] = token
+
+    # ── 获取 Chat ID ─────────────────────────────────────────────────────────
+    existing_ids = env.get("TELEGRAM_ALLOWED_IDS", "")
+    ids_str = _ask(
+        "TELEGRAM_ALLOWED_IDS (你的 Chat ID，多个用英文逗号分隔)",
+        default=existing_ids if existing_ids != "123456789" else "",
+    )
+    if ids_str:
+        env["TELEGRAM_ALLOWED_IDS"] = ids_str
+        _ok(f"已允许 Chat ID: {ids_str}")
+    else:
+        _warn("未填写 Chat ID，Bot 将拒绝所有消息")
+        _info("获取方法：Telegram 搜索 @userinfobot 发送任意消息即可")
+
+    _ok("Telegram Bot 配置完成")
+    if _rich:
+        console.print(
+            "\n  [dim]启动 daemon 后向你的 Bot 发 /start 即可开始使用[/dim]\n"
+            "  [cyan]python3 aria_daemon.py[/cyan]\n"
+        )
+    else:
+        print("\n  启动: python3 aria_daemon.py")
+
+
+def _verify_telegram_token(token: str) -> dict | None:
+    """Call getMe to verify a Telegram bot token. Returns bot info dict or None."""
+    import urllib.request as _req
+    import urllib.error as _err
+    try:
+        url = f"https://api.telegram.org/bot{token}/getMe"
+        with _req.urlopen(url, timeout=5) as resp:
+            import json as _j
+            data = _j.loads(resp.read())
+            if data.get("ok"):
+                return data.get("result", {})
+    except (_err.URLError, Exception):
+        pass
+    return None
+
+
+# ── Step 5: Write config & finalize ──────────────────────────────────────────
+
+def _write_aria_config(env: dict[str, str]) -> None:
+    """Write ~/.aria/config.json for aria_cli.py to read."""
+    cfg: dict = {}
+    if _CFG_FILE.exists():
+        try:
+            cfg = json.loads(_CFG_FILE.read_text())
+        except Exception:
+            cfg = {}
+
+    model = env.get("ARIA_DEFAULT_MODEL", "")
+    if model:
+        cfg["default_model"] = model
+        cfg["model"] = model  # also write as "model" for aria_cli.py
+
+    # Persist detected system language so UI renders correctly from first launch
+    cfg["ui_lang"] = _SYS_LANG
+
+    _ARIA_DIR.mkdir(parents=True, exist_ok=True)
+    _CFG_FILE.write_text(json.dumps(cfg, indent=2, ensure_ascii=False))
+
+
+def _print_summary(env: dict[str, str]) -> None:
+    _section("配置完成")
+    if _rich:
+        t = Table(box=rbox.SIMPLE, show_header=False)
+        t.add_column("项目", style="dim", width=22)
+        t.add_column("值",   style="cyan")
+
+        def _mask(v: str) -> str:
+            return v[:6] + "…" + v[-4:] if len(v) > 12 else "***"
+
+        t.add_row("默认模型",   env.get("ARIA_DEFAULT_MODEL", "—"))
+        t.add_row("飞书模式",   env.get("ARIA_RELAY_MODE", "—"))
+        if env.get("ARIA_RELAY_CLIENT_ID"):
+            t.add_row("中继 Client ID", env["ARIA_RELAY_CLIENT_ID"])
+        if env.get("FEISHU_APP_ID"):
+            t.add_row("飞书 App ID",   env["FEISHU_APP_ID"])
+        tg_token = env.get("TELEGRAM_BOT_TOKEN", "")
+        if tg_token and tg_token != "your_bot_token_here":
+            t.add_row("Telegram Bot",  _mask(tg_token))
+        if env.get("TELEGRAM_ALLOWED_IDS"):
+            t.add_row("Telegram IDs",  env["TELEGRAM_ALLOWED_IDS"])
+        if env.get("ANTHROPIC_API_KEY"):
+            t.add_row("Anthropic Key", _mask(env["ANTHROPIC_API_KEY"]))
+        if env.get("OPENAI_API_KEY"):
+            t.add_row("OpenAI Key",    _mask(env["OPENAI_API_KEY"]))
+        console.print(t)
+
+        console.print(Panel(
+            "[bold green]Aria Code 配置成功！[/bold green]\n\n"
+            "启动方式：\n"
+            "  [cyan]python3 aria_cli.py[/cyan]              # 交互终端\n"
+            "  [cyan]python3 aria_daemon.py --install[/cyan] # 安装为系统服务（开机自启）\n"
+            "  [cyan]python3 aria_daemon.py[/cyan]           # 前台运行 daemon\n\n"
+            "[dim]配置文件: ~/.aria/.env[/dim]",
+            border_style="green",
+        ))
+    else:
+        print("\n配置完成！")
+        print(f"默认模型: {env.get('ARIA_DEFAULT_MODEL', '—')}")
+        print(f"飞书模式: {env.get('ARIA_RELAY_MODE', '—')}")
+        print("\n启动: python3 aria_daemon.py")
+
+
+# ── Main ─────────────────────────────────────────────────────────────────────
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description="Aria Code Setup Wizard / 配置向导")
+    parser.add_argument("--model",    action="store_true", help="Model setup only / 仅配置模型")
+    parser.add_argument("--feishu",   action="store_true", help="Feishu only / 仅配置飞书")
+    parser.add_argument("--telegram", action="store_true", help="Telegram only / 仅配置 Telegram")
+    parser.add_argument("--keys",     action="store_true", help="API keys only / 仅配置 API Key")
+    args = parser.parse_args()
+
+    # Show detected language
+    _info(_wz("lang_detected"))
+
+    if _SYS_LANG == "zh":
+        _intro = (
+            "[bold cyan]Aria Code[/bold cyan] [dim]— AI 金融终端[/dim]\n\n"
+            "本向导将帮助你完成首次配置：\n"
+            "  • 本地 AI 模型（Ollama）\n"
+            "  • API 密钥（Claude / GPT-4，可选）\n"
+            "  • 飞书机器人连接\n"
+            "  • Telegram Bot 接入"
+        )
+        _title = "🔧 首次配置向导"
+        _plain_intro = "\n=== Aria Code 配置向导 ===\n"
+    else:
+        _intro = (
+            "[bold cyan]Aria Code[/bold cyan] [dim]— AI Finance Terminal[/dim]\n\n"
+            "This wizard will help you set up:\n"
+            "  • Local AI model (Ollama)\n"
+            "  • API keys (Claude / GPT-4 — optional)\n"
+            "  • Feishu / Lark bot connection\n"
+            "  • Telegram Bot integration"
+        )
+        _title = "🔧 First-Run Setup"
+        _plain_intro = "\n=== Aria Code Setup Wizard ===\n"
+
+    if _rich:
+        console.print(Panel(_intro, title=_title, border_style="cyan"))
+    else:
+        print(_plain_intro)
+
+    env = _load_env()
+    run_all = not (args.model or args.feishu or args.telegram or args.keys)
+
+    if run_all or args.model:
+        setup_model(env)
+
+    if run_all or args.keys:
+        setup_api_keys(env)
+
+    if run_all or args.feishu:
+        setup_feishu(env)
+
+    if run_all or args.telegram:
+        setup_telegram(env)
+
+    _save_env(env)
+    _write_aria_config(env)
+    _ok(f"配置已保存到 {_ENV_FILE}")
+
+    if run_all:
+        _print_summary(env)
+
+    if run_all and _confirm("\n是否立即启动 Aria Daemon？", default=True):
+        os.execv(sys.executable, [sys.executable, str(Path(__file__).parent / "aria_daemon.py")])
+
+
+if __name__ == "__main__":
+    main()