🔄 卡若AI 同步 2026-03-15 19:47 | 更新：水溪整理归档、卡木、火炬、总索引与入口、运营中枢工作台 | 排除 >20MB: 11 个

02_卡人（水）/水溪_整理归档/经验库/待沉淀/2026-03-15_全站埋点统计标准_Soul项目沉淀.md

@@ -0,0 +1,74 @@
+# 全站埋点统计标准（Soul 创业实验项目沉淀）
+
+> 日期：2026-03-15  
+> 来源：Soul 创业实验项目 · 数据统计迭代  
+> 已写入：全栈开发 SKILL v2.5 § 1.10
+
+---
+
+## 背景
+
+Soul 创业实验项目在管理后台需要「分类标签点击统计」面板，但发现小程序端没有埋点数据上报。从零实施全站埋点，覆盖 9 个小程序页面、后端聚合 API、管理后台可视化面板，前后共 3 小时完成。
+
+**教训**：功能上线后再补埋点，已丢失大量初期用户行为数据。**以后功能和埋点必须同时上线。**
+
+---
+
+## 标准三层架构
+
+```
+前端（小程序/Web）            后端 API                     管理后台
+trackClick(module,        →  POST /api/{平台}/track    →  GET /api/admin/track/stats
+  action, target, extra)      存入 user_tracks 表           按 module/action/时间段聚合
+```
+
+### 1. 前端：trackClick 工具
+
+每个平台（小程序、Web、App）封装一个 `trackClick(module, action, target, extra?)` 函数：
+
+- **module**：所属模块（home / chapters / read / my / vip / wallet / match / referral / search）
+- **action**：标准动词（btn_click / page_view / tab_click / nav_click / share）
+- **target**：具体按钮/链接名（如「购买VIP」「充值」「阅读第3章」）
+- **extra**：可选扩展字段
+
+静默上报，不弹窗不阻塞用户操作。
+
+### 2. 后端：track API + user_tracks 表
+
+| 字段 | 类型 | 说明 |
+|:---|:---|:---|
+| id | string/UUID | 主键 |
+| user_id | string | 用户 ID |
+| action | string | 动作类型 |
+| target | string | 具体目标 |
+| extra_data | JSONB | 扩展信息，含 module、page |
+| created_at | timestamp | 自动时间戳 |
+
+注意 `extra_data` 要正确 Marshal 为 JSON 存入 JSONB 字段（Go 端踩过坑：忘记 marshal body.ExtraData 导致 null）。
+
+### 3. 管理后台：聚合面板
+
+- 后端 `GET /api/admin/track/stats?period=today|week|month|all` 按 action+target+extra_data 聚合并分组
+- 前端面板按 module 分组展示，支持时间段切换，30 秒自动刷新
+
+---
+
+## 埋点接入检查清单
+
+每开发一个新功能/新页面，按此清单打勾：
+
+- [ ] 页面所有可点击按钮/标签都已调用 trackClick
+- [ ] module 参数使用统一命名
+- [ ] action 参数使用标准动词
+- [ ] target 参数能区分具体按钮
+- [ ] 后端 track API 已注册路由且能正确存储 extra_data
+- [ ] 管理后台能展示该模块数据
+
+---
+
+## 实际代码参考
+
+- **小程序 trackClick**：`miniprogram/utils/trackClick.js`
+- **后端 track handler**：`soul-api/internal/handler/admin_track.go`（聚合统计）、`soul-api/internal/handler/user.go`（UserTrackPost 存储）
+- **管理后台面板**：`soul-admin/src/pages/dashboard/DashboardPage.tsx`（分类标签点击统计 Card）
+- **数据模型**：`soul-api/internal/model/user_track.go`

03_卡木（木）/木根_逆向分析/全网AI自动注册/SKILL.md

@@ -1,14 +1,14 @@
 ---
-name: 全网API自动注册
-description: 自动注册全网各类 AI/开发 API 免费账号，提取 API Key/Token 并统一管理
-triggers: API注册、自动注册、批量注册、API Key、注册账号、免费API、API池、key池、自动开号
+name: 全网AI自动注册
+description: 自动注册全网各类 AI API 免费账号，提取 API Key/Token 并统一管理
+triggers: AI注册、自动注册、批量注册、API Key、注册账号、免费API、API池、key池、自动开号、Gemini注册
 owner: 木根
 group: 木（卡木）
 version: "1.0"
 updated: "2026-03-15"
 ---
 
-# 全网API自动注册
+# 全网AI自动注册
 
 ## 能做什么（Capabilities）
 

03_卡木（木）/木根_逆向分析/全网AI自动注册/脚本/gemini_key_creator.py

@@ -0,0 +1,320 @@
+#!/usr/bin/env python3
+"""
+Gemini API Key 创建器（纯命令行 + 系统浏览器 OAuth 授权）
+不需要打开额外浏览器，系统浏览器走 Clash 代理完成 Google 登录授权。
+自动创建指定数量的 GCP 项目并生成 Gemini API Key。
+
+用法: python3 gemini_key_creator.py --count 3
+"""
+
+import argparse
+import http.server
+import json
+import os
+import random
+import string
+import sys
+import threading
+import time
+import urllib.parse
+import webbrowser
+from pathlib import Path
+
+import httpx
+
+PROXY = "http://127.0.0.1:7897"
+
+# Google OAuth2 客户端（使用 Google Cloud SDK 的公开 client）
+CLIENT_ID = "764086051850-6qr4p6gpi6hn506pt8ejuq83di341hur.apps.googleusercontent.com"
+CLIENT_SECRET = "d-FL95Q19q7MQmFpd7hHD0Ty"
+REDIRECT_PORT = 18457
+REDIRECT_URI = f"http://localhost:{REDIRECT_PORT}"
+SCOPES = [
+    "https://www.googleapis.com/auth/cloud-platform",
+    "https://www.googleapis.com/auth/generative-language",
+]
+
+TOKEN_URL = "https://oauth2.googleapis.com/token"
+PROJECTS_URL = "https://cloudresourcemanager.googleapis.com/v1/projects"
+SERVICES_URL = "https://serviceusage.googleapis.com/v1/projects/{project_id}/services/generativelanguage.googleapis.com:enable"
+APIKEYS_URL = "https://apikeys.googleapis.com/v2/projects/{project_id}/locations/global/keys"
+
+DB_PATH = Path(__file__).parent / "accounts.db"
+JSON_DIR = Path(__file__).parent / "tokens"
+
+auth_code_result = {"code": None}
+
+
+class OAuthCallbackHandler(http.server.BaseHTTPRequestHandler):
+    def do_GET(self):
+        parsed = urllib.parse.urlparse(self.path)
+        qs = urllib.parse.parse_qs(parsed.query)
+        code = qs.get("code", [None])[0]
+        if code:
+            auth_code_result["code"] = code
+            self.send_response(200)
+            self.send_header("Content-Type", "text/html; charset=utf-8")
+            self.end_headers()
+            self.wfile.write(b"<html><body><h2>&#10004; Google OAuth OK</h2><p>You can close this tab now.</p></body></html>")
+        else:
+            error = qs.get("error", ["unknown"])[0]
+            self.send_response(400)
+            self.send_header("Content-Type", "text/html")
+            self.end_headers()
+            self.wfile.write(f"<html><body><h2>Error: {error}</h2></body></html>".encode())
+
+    def log_message(self, format, *args):
+        pass
+
+
+def get_oauth_token():
+    """OAuth 授权码流程：系统浏览器授权 → 获取 access_token"""
+    server = http.server.HTTPServer(("localhost", REDIRECT_PORT), OAuthCallbackHandler)
+    server.timeout = 300
+    thread = threading.Thread(target=server.handle_request, daemon=True)
+    thread.start()
+
+    auth_params = urllib.parse.urlencode({
+        "client_id": CLIENT_ID,
+        "redirect_uri": REDIRECT_URI,
+        "response_type": "code",
+        "scope": " ".join(SCOPES),
+        "access_type": "offline",
+        "prompt": "consent",
+    })
+    auth_url = f"https://accounts.google.com/o/oauth2/v2/auth?{auth_params}"
+
+    print(f"\n🔐 正在打开系统浏览器进行 Google 授权...")
+    print(f"   （系统浏览器走 Clash 代理，无需额外操作）\n")
+    webbrowser.open(auth_url)
+
+    print("⏳ 等待授权回调...")
+    thread.join(timeout=300)
+    server.server_close()
+
+    code = auth_code_result["code"]
+    if not code:
+        print("❌ 授权超时或失败")
+        sys.exit(1)
+
+    print("✅ 授权码获取成功，兑换 Token...")
+    with httpx.Client(proxy=PROXY, timeout=20) as client:
+        resp = client.post(TOKEN_URL, data={
+            "code": code,
+            "client_id": CLIENT_ID,
+            "client_secret": CLIENT_SECRET,
+            "redirect_uri": REDIRECT_URI,
+            "grant_type": "authorization_code",
+        })
+        if resp.status_code != 200:
+            print(f"❌ Token 兑换失败: {resp.status_code} {resp.text[:200]}")
+            sys.exit(1)
+        tokens = resp.json()
+        print(f"✅ Access Token 获取成功!")
+        return tokens["access_token"]
+
+
+def create_project(token, project_id):
+    """创建 GCP 项目"""
+    with httpx.Client(proxy=PROXY, timeout=30) as client:
+        resp = client.post(
+            PROJECTS_URL,
+            headers={"Authorization": f"Bearer {token}"},
+            json={"projectId": project_id, "name": project_id},
+        )
+        if resp.status_code in (200, 409):
+            if resp.status_code == 409:
+                print(f"  项目 {project_id} 已存在，继续使用")
+            return True
+        print(f"  创建项目失败: {resp.status_code} {resp.text[:200]}")
+        return False
+
+
+def enable_gemini_api(token, project_id):
+    """启用 Generative Language API"""
+    url = SERVICES_URL.format(project_id=project_id)
+    with httpx.Client(proxy=PROXY, timeout=30) as client:
+        resp = client.post(
+            url,
+            headers={"Authorization": f"Bearer {token}"},
+        )
+        if resp.status_code in (200, 409):
+            return True
+        print(f"  启用 API 失败: {resp.status_code} {resp.text[:200]}")
+        return False
+
+
+def wait_for_operation(token, operation_name, max_wait=60):
+    """等待长期运行操作完成"""
+    url = f"https://serviceusage.googleapis.com/v1/{operation_name}"
+    with httpx.Client(proxy=PROXY, timeout=15) as client:
+        for _ in range(max_wait // 3):
+            resp = client.get(url, headers={"Authorization": f"Bearer {token}"})
+            if resp.status_code == 200:
+                data = resp.json()
+                if data.get("done"):
+                    return True
+            time.sleep(3)
+    return True
+
+
+def create_api_key(token, project_id, display_name):
+    """创建 API Key 并限制为 Generative Language API"""
+    url = APIKEYS_URL.format(project_id=project_id)
+    with httpx.Client(proxy=PROXY, timeout=30) as client:
+        resp = client.post(
+            url,
+            headers={"Authorization": f"Bearer {token}"},
+            json={
+                "displayName": display_name,
+                "restrictions": {
+                    "apiTargets": [{"service": "generativelanguage.googleapis.com"}]
+                },
+            },
+        )
+        if resp.status_code == 200:
+            data = resp.json()
+            operation = data.get("name", "")
+            if "operations/" in operation:
+                time.sleep(3)
+                op_resp = client.get(
+                    f"https://apikeys.googleapis.com/v2/{operation}",
+                    headers={"Authorization": f"Bearer {token}"},
+                )
+                if op_resp.status_code == 200:
+                    op_data = op_resp.json()
+                    if op_data.get("done"):
+                        key_data = op_data.get("response", {})
+                        return key_data.get("keyString", "")
+                    time.sleep(5)
+                    op_resp2 = client.get(
+                        f"https://apikeys.googleapis.com/v2/{operation}",
+                        headers={"Authorization": f"Bearer {token}"},
+                    )
+                    if op_resp2.status_code == 200:
+                        return op_resp2.json().get("response", {}).get("keyString", "")
+            return data.get("keyString", "")
+        print(f"  创建 Key 失败: {resp.status_code} {resp.text[:300]}")
+        return None
+
+
+def list_existing_keys(token, project_id):
+    """列出已有的 API Keys"""
+    url = APIKEYS_URL.format(project_id=project_id)
+    with httpx.Client(proxy=PROXY, timeout=15) as client:
+        resp = client.get(url, headers={"Authorization": f"Bearer {token}"})
+        if resp.status_code == 200:
+            return resp.json().get("keys", [])
+    return []
+
+
+def save_key_to_db(api_key, project_id, idx):
+    """保存到 SQLite"""
+    import sqlite3
+    JSON_DIR.mkdir(parents=True, exist_ok=True)
+
+    with sqlite3.connect(str(DB_PATH)) as conn:
+        conn.execute("""
+            CREATE TABLE IF NOT EXISTS accounts (
+                id INTEGER PRIMARY KEY AUTOINCREMENT,
+                provider TEXT NOT NULL,
+                email TEXT NOT NULL,
+                password TEXT DEFAULT '',
+                api_key TEXT DEFAULT '',
+                access_token TEXT DEFAULT '',
+                refresh_token TEXT DEFAULT '',
+                account_id TEXT DEFAULT '',
+                name TEXT DEFAULT '',
+                extra TEXT DEFAULT '{}',
+                registered_at TEXT NOT NULL,
+                status TEXT DEFAULT 'active',
+                UNIQUE(provider, email)
+            )
+        """)
+        from datetime import datetime, timezone
+        now = datetime.now(timezone.utc).isoformat()
+        conn.execute("""
+            INSERT OR REPLACE INTO accounts
+            (provider, email, api_key, name, extra, registered_at, status)
+            VALUES (?, ?, ?, ?, ?, ?, 'active')
+        """, ("gemini", f"gemini_project_{idx}@google.com", api_key,
+              project_id, json.dumps({"project": project_id}), now))
+        conn.commit()
+
+    json_file = JSON_DIR / f"gemini_{project_id}.json"
+    json_file.write_text(json.dumps({
+        "provider": "gemini",
+        "project": project_id,
+        "api_key": api_key,
+    }, indent=2))
+    print(f"  💾 已保存: DB + {json_file.name}")
+
+
+def main():
+    global PROXY
+    parser = argparse.ArgumentParser(description="Gemini API Key 创建器")
+    parser.add_argument("--count", "-n", type=int, default=3, help="创建 Key 数量")
+    parser.add_argument("--proxy", default=PROXY, help="代理地址")
+    args = parser.parse_args()
+
+    PROXY = args.proxy
+
+    print("=" * 60)
+    print(f"🔑 Gemini API Key 创建器")
+    print(f"   目标: 创建 {args.count} 个 Gemini API Key")
+    print(f"   代理: {PROXY}")
+    print("=" * 60)
+
+    token = get_oauth_token()
+
+    keys_created = []
+    for i in range(1, args.count + 1):
+        suffix = "".join(random.choices(string.ascii_lowercase + string.digits, k=6))
+        project_id = f"gemini-auto-{suffix}"
+        display_name = f"gemini-key-{i}"
+
+        print(f"\n--- 创建第 {i}/{args.count} 个 Key ---")
+
+        print(f"  [1/3] 创建项目: {project_id}")
+        if not create_project(token, project_id):
+            print(f"  ⚠️  跳过此 Key")
+            continue
+
+        print(f"  [2/3] 启用 Generative Language API")
+        time.sleep(2)
+        enable_gemini_api(token, project_id)
+        time.sleep(5)
+
+        print(f"  [3/3] 创建 API Key: {display_name}")
+        api_key = create_api_key(token, project_id, display_name)
+
+        if api_key:
+            print(f"  ✅ Key {i}: {api_key}")
+            keys_created.append({"key": api_key, "project": project_id})
+            save_key_to_db(api_key, project_id, i)
+        else:
+            print(f"  ❌ Key 创建失败，尝试列出已有 Key...")
+            existing = list_existing_keys(token, project_id)
+            for k in existing:
+                ks = k.get("keyString")
+                if ks:
+                    print(f"  ✅ 找到已有 Key: {ks}")
+                    keys_created.append({"key": ks, "project": project_id})
+                    save_key_to_db(ks, project_id, i)
+                    break
+
+    print(f"\n{'='*60}")
+    print(f"🎉 完成! 创建了 {len(keys_created)}/{args.count} 个 Gemini API Key")
+    for idx, kd in enumerate(keys_created, 1):
+        print(f"  Key {idx}: {kd['key']}")
+        print(f"         Project: {kd['project']}")
+    if keys_created:
+        print(f"\n测试命令:")
+        k = keys_created[0]["key"]
+        print(f'  curl -x {PROXY} "https://generativelanguage.googleapis.com/v1beta/models?key={k}"')
+    print(f"\n存储: {DB_PATH} + {JSON_DIR}/")
+    print(f"{'='*60}")
+
+
+if __name__ == "__main__":
+    main()

03_卡木（木）/木根_逆向分析/全网AI自动注册/脚本/google_device_auth.py

@@ -0,0 +1,157 @@
+#!/usr/bin/env python3
+"""Google Device Flow 授权 + 创建 3 个 Gemini API Key"""
+
+import httpx, time, json, sys, random, string, sqlite3
+from pathlib import Path
+from datetime import datetime, timezone
+
+PROXY = "http://127.0.0.1:7897"
+CLIENT_ID = "764086051850-6qr4p6gpi6hn506pt8ejuq83di341hur.apps.googleusercontent.com"
+CLIENT_SECRET = "d-FL95Q19q7MQmFpd7hHD0Ty"
+SCOPES = "https://www.googleapis.com/auth/cloud-platform"
+DB_PATH = Path(__file__).parent / "accounts.db"
+JSON_DIR = Path(__file__).parent / "tokens"
+COUNT = int(sys.argv[1]) if len(sys.argv) > 1 else 3
+
+
+def save_key(api_key, project_id, idx):
+    JSON_DIR.mkdir(parents=True, exist_ok=True)
+    with sqlite3.connect(str(DB_PATH)) as conn:
+        conn.execute("""CREATE TABLE IF NOT EXISTS accounts (
+            id INTEGER PRIMARY KEY AUTOINCREMENT, provider TEXT NOT NULL,
+            email TEXT NOT NULL, password TEXT DEFAULT '', api_key TEXT DEFAULT '',
+            access_token TEXT DEFAULT '', refresh_token TEXT DEFAULT '',
+            account_id TEXT DEFAULT '', name TEXT DEFAULT '', extra TEXT DEFAULT '{}',
+            registered_at TEXT NOT NULL, status TEXT DEFAULT 'active',
+            UNIQUE(provider, email))""")
+        now = datetime.now(timezone.utc).isoformat()
+        conn.execute("INSERT OR REPLACE INTO accounts (provider,email,api_key,name,extra,registered_at,status) VALUES (?,?,?,?,?,?,'active')",
+            ("gemini", f"gemini_{project_id}@gcp", api_key, project_id, json.dumps({"project": project_id}), now))
+        conn.commit()
+    (JSON_DIR / f"gemini_{project_id}.json").write_text(json.dumps({"provider":"gemini","project":project_id,"api_key":api_key}, indent=2))
+
+
+def main():
+    sep = "=" * 50
+    print(sep)
+    print("Gemini API Key Creator (Device Flow)")
+    print(f"Target: {COUNT} keys | Proxy: {PROXY}")
+    print(sep)
+
+    # Step 1: Device Flow
+    print("\n[Step 1] Google Device Flow...")
+    with httpx.Client(proxy=PROXY, timeout=15) as c:
+        r = c.post("https://oauth2.googleapis.com/device/code", data={"client_id": CLIENT_ID, "scope": SCOPES})
+        print(f"Status: {r.status_code}")
+        if r.status_code != 200:
+            print(f"Failed: {r.text[:300]}")
+            return
+
+        d = r.json()
+        user_code = d["user_code"]
+        verify_url = d["verification_url"]
+        device_code = d["device_code"]
+        interval = d.get("interval", 5)
+
+        print(f"\n{sep}")
+        print(f"Open in ANY browser/phone: {verify_url}")
+        print(f"Enter code: {user_code}")
+        print(f"{sep}\n")
+        print("Waiting for authorization...")
+
+        import webbrowser
+        webbrowser.open(verify_url)
+
+        token = None
+        for i in range(120):
+            time.sleep(interval)
+            tr = c.post("https://oauth2.googleapis.com/token", data={
+                "client_id": CLIENT_ID, "client_secret": CLIENT_SECRET,
+                "device_code": device_code, "grant_type": "urn:ietf:params:oauth:grant-type:device_code"})
+            td = tr.json()
+            if "access_token" in td:
+                token = td["access_token"]
+                print(f"\nAuthorized! Token: {token[:25]}...")
+                break
+            err = td.get("error")
+            if err == "authorization_pending":
+                if i % 12 == 0 and i > 0:
+                    print(f"  Still waiting... ({i*interval}s)")
+            elif err == "slow_down":
+                time.sleep(5)
+            else:
+                print(f"Error: {td}")
+                return
+
+        if not token:
+            print("Timeout!")
+            return
+
+    # Step 2: Create projects and keys
+    print(f"\n[Step 2] Creating {COUNT} Gemini API Keys...")
+    keys = []
+    with httpx.Client(proxy=PROXY, timeout=30) as c:
+        headers = {"Authorization": f"Bearer {token}"}
+        for i in range(1, COUNT + 1):
+            sfx = "".join(random.choices(string.ascii_lowercase + string.digits, k=6))
+            pid = f"gemini-auto-{sfx}"
+            print(f"\n  Key {i}/{COUNT}: project={pid}")
+
+            # Create project
+            pr = c.post("https://cloudresourcemanager.googleapis.com/v1/projects",
+                        headers=headers, json={"projectId": pid, "name": pid})
+            print(f"    Create project: {pr.status_code}")
+            if pr.status_code not in (200, 409):
+                print(f"    Failed: {pr.text[:200]}")
+                continue
+
+            time.sleep(3)
+
+            # Enable API
+            er = c.post(f"https://serviceusage.googleapis.com/v1/projects/{pid}/services/generativelanguage.googleapis.com:enable",
+                        headers=headers)
+            print(f"    Enable API: {er.status_code}")
+            time.sleep(5)
+
+            # Create API key
+            kr = c.post(f"https://apikeys.googleapis.com/v2/projects/{pid}/locations/global/keys",
+                        headers=headers, json={"displayName": f"gemini-key-{i}",
+                            "restrictions": {"apiTargets": [{"service": "generativelanguage.googleapis.com"}]}})
+            print(f"    Create key: {kr.status_code}")
+
+            if kr.status_code == 200:
+                kd = kr.json()
+                op_name = kd.get("name", "")
+                if "operations/" in op_name:
+                    for _ in range(10):
+                        time.sleep(3)
+                        or2 = c.get(f"https://apikeys.googleapis.com/v2/{op_name}", headers=headers)
+                        if or2.status_code == 200:
+                            od = or2.json()
+                            if od.get("done"):
+                                ks = od.get("response", {}).get("keyString", "")
+                                if ks:
+                                    print(f"    KEY: {ks}")
+                                    keys.append({"key": ks, "project": pid})
+                                    save_key(ks, pid, i)
+                                break
+                else:
+                    ks = kd.get("keyString", "")
+                    if ks:
+                        print(f"    KEY: {ks}")
+                        keys.append({"key": ks, "project": pid})
+                        save_key(ks, pid, i)
+            else:
+                print(f"    Failed: {kr.text[:300]}")
+
+    print(f"\n{sep}")
+    print(f"Done! {len(keys)}/{COUNT} keys created")
+    for idx, k in enumerate(keys, 1):
+        print(f"  {idx}. {k['key']}  (project: {k['project']})")
+    if keys:
+        print(f"\nTest: curl -x {PROXY} 'https://generativelanguage.googleapis.com/v1beta/models?key={keys[0]['key']}'")
+    print(sep)
+
+
+if __name__ == "__main__":
+    main()

03_卡木（木）/木根_逆向分析/全网AI自动注册/脚本/register_gemini_keys.py

@@ -0,0 +1,271 @@
+#!/usr/bin/env python3
+"""
+Gemini API Key 自动注册脚本
+1. 通过 Clash 代理打开 Google AI Studio（非无头模式，你手动登录 Google）
+2. 登录完成后脚本自动创建指定数量的 API Key
+3. Key 存入本地 SQLite + JSON
+
+用法: python3 register_gemini_keys.py --count 3
+"""
+
+import argparse
+import json
+import os
+import re
+import sys
+import time
+from pathlib import Path
+from datetime import datetime, timezone
+
+PROXY_SERVER = "http://127.0.0.1:7897"
+AI_STUDIO_URL = "https://aistudio.google.com/app/apikey"
+DB_PATH = Path(__file__).parent / "accounts.db"
+JSON_DIR = Path(__file__).parent / "tokens"
+
+
+def save_key(provider, email, api_key, project_name=""):
+    """保存到 SQLite + JSON"""
+    import sqlite3
+    JSON_DIR.mkdir(parents=True, exist_ok=True)
+
+    with sqlite3.connect(str(DB_PATH)) as conn:
+        conn.execute("""
+            CREATE TABLE IF NOT EXISTS accounts (
+                id INTEGER PRIMARY KEY AUTOINCREMENT,
+                provider TEXT NOT NULL,
+                email TEXT NOT NULL,
+                password TEXT DEFAULT '',
+                api_key TEXT DEFAULT '',
+                access_token TEXT DEFAULT '',
+                refresh_token TEXT DEFAULT '',
+                account_id TEXT DEFAULT '',
+                name TEXT DEFAULT '',
+                extra TEXT DEFAULT '{}',
+                registered_at TEXT NOT NULL,
+                status TEXT DEFAULT 'active',
+                UNIQUE(provider, email)
+            )
+        """)
+        now = datetime.now(timezone.utc).isoformat()
+        conn.execute("""
+            INSERT OR REPLACE INTO accounts
+            (provider, email, api_key, name, extra, registered_at, status)
+            VALUES (?, ?, ?, ?, ?, ?, 'active')
+        """, (provider, email, api_key, project_name, json.dumps({"project": project_name}), now))
+        conn.commit()
+
+    safe_email = email.replace("@", "_at_")
+    json_file = JSON_DIR / f"gemini_{safe_email}_{int(time.time())}.json"
+    json_file.write_text(json.dumps({
+        "provider": "gemini",
+        "email": email,
+        "api_key": api_key,
+        "project": project_name,
+        "registered_at": datetime.now(timezone.utc).isoformat(),
+    }, indent=2, ensure_ascii=False))
+    print(f"  [保存] {json_file.name}")
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Gemini API Key 自动注册")
+    parser.add_argument("--count", "-n", type=int, default=3, help="要创建的 Key 数量")
+    parser.add_argument("--proxy", default=PROXY_SERVER, help="代理地址")
+    args = parser.parse_args()
+
+    from playwright.sync_api import sync_playwright
+
+    print(f"🚀 启动浏览器（代理: {args.proxy}）")
+    print(f"📋 目标: 创建 {args.count} 个 Gemini API Key\n")
+
+    pw = sync_playwright().start()
+    browser = pw.chromium.launch(
+        headless=False,
+        proxy={"server": args.proxy},
+        args=["--window-size=1280,900"],
+    )
+    context = browser.new_context(
+        viewport={"width": 1280, "height": 900},
+        locale="en-US",
+    )
+    page = context.new_page()
+
+    print("📂 打开 Google AI Studio...")
+    page.goto(AI_STUDIO_URL, timeout=30000)
+    time.sleep(2)
+
+    if "accounts.google.com" in page.url:
+        print("\n" + "=" * 60)
+        print("⚠️  请在弹出的浏览器窗口中登录你的 Google 账号")
+        print("   登录完成后脚本会自动继续...")
+        print("=" * 60 + "\n")
+
+        for i in range(300):
+            time.sleep(2)
+            current = page.url
+            # 必须是真正到达 aistudio，而非 accounts.google.com 里的 continue 参数
+            if current.startswith("https://aistudio.google.com"):
+                print("✅ 登录成功！已进入 AI Studio")
+                break
+            if i % 15 == 0 and i > 0:
+                print(f"  等待登录中... ({i*2}秒)")
+        else:
+            print("❌ 登录超时（10分钟），请重试")
+            browser.close()
+            pw.stop()
+            return
+
+    time.sleep(3)
+    print(f"\n当前页面: {page.url}")
+
+    if "/apikey" not in page.url:
+        page.goto(AI_STUDIO_URL, timeout=20000)
+        time.sleep(3)
+
+    keys_created = []
+
+    for i in range(args.count):
+        print(f"\n--- 创建第 {i+1}/{args.count} 个 Key ---")
+
+        try:
+            page.wait_for_load_state("networkidle", timeout=10000)
+        except Exception:
+            pass
+
+        create_btn = None
+        for selector in [
+            "button:has-text('Create API key')",
+            "button:has-text('Create API Key')",
+            "button:has-text('创建 API 密钥')",
+            "[aria-label='Create API key']",
+            "button:has-text('Get API key')",
+            "button:has-text('获取 API 密钥')",
+        ]:
+            try:
+                btn = page.locator(selector).first
+                if btn.is_visible(timeout=3000):
+                    create_btn = btn
+                    break
+            except Exception:
+                continue
+
+        if not create_btn:
+            print("  ⚠️  未找到 Create API Key 按钮，尝试截图诊断...")
+            page.screenshot(path=str(JSON_DIR / f"debug_step_{i+1}.png"))
+            snapshot = page.content()
+            if "Create" in snapshot or "创建" in snapshot:
+                print("  页面包含创建按钮文字，尝试通用点击...")
+                try:
+                    page.locator("button").filter(has_text=re.compile(r"Create|创建|Get|获取")).first.click()
+                except Exception as e:
+                    print(f"  点击失败: {e}")
+                    continue
+            else:
+                print(f"  页面 URL: {page.url}")
+                continue
+        else:
+            create_btn.click()
+
+        time.sleep(2)
+
+        new_project_btn = None
+        for selector in [
+            "button:has-text('Create API key in new project')",
+            "button:has-text('在新项目中创建 API 密钥')",
+            "text='Create API key in new project'",
+            "button:has-text('new project')",
+        ]:
+            try:
+                btn = page.locator(selector).first
+                if btn.is_visible(timeout=3000):
+                    new_project_btn = btn
+                    break
+            except Exception:
+                continue
+
+        if new_project_btn:
+            new_project_btn.click()
+            print("  点击了 'Create in new project'")
+        else:
+            existing_btns = page.locator("button:has-text('Create')").all()
+            for btn in existing_btns:
+                try:
+                    if btn.is_visible():
+                        btn.click()
+                        break
+                except Exception:
+                    continue
+
+        time.sleep(5)
+
+        api_key = None
+        for attempt in range(10):
+            page_text = page.content()
+            matches = re.findall(r'AIzaSy[A-Za-z0-9_-]{33}', page_text)
+            if matches:
+                api_key = matches[0]
+                break
+
+            try:
+                code_el = page.locator("code, .api-key, [data-testid*='key'], pre").first
+                if code_el.is_visible(timeout=1000):
+                    text = code_el.inner_text()
+                    m = re.search(r'AIzaSy[A-Za-z0-9_-]{33}', text)
+                    if m:
+                        api_key = m.group(0)
+                        break
+            except Exception:
+                pass
+
+            copy_btns = page.locator("button:has-text('Copy')").all()
+            if not copy_btns:
+                copy_btns = page.locator("button:has-text('复制')").all()
+            for btn in copy_btns:
+                try:
+                    if btn.is_visible():
+                        btn.click()
+                        time.sleep(0.5)
+                        break
+                except Exception:
+                    pass
+
+            time.sleep(1)
+            if attempt % 3 == 2:
+                print(f"  等待 Key 生成... (尝试 {attempt+1})")
+
+        if api_key:
+            print(f"  ✅ Key {i+1}: {api_key[:20]}...")
+            keys_created.append(api_key)
+            save_key("gemini", f"gemini_project_{i+1}@google.com", api_key, f"project_{i+1}")
+        else:
+            print(f"  ❌ 未能提取到 API Key，请检查浏览器窗口")
+            page.screenshot(path=str(JSON_DIR / f"failed_key_{i+1}.png"))
+
+        close_btns = page.locator("button:has-text('Done'), button:has-text('Close'), button:has-text('完成'), button:has-text('关闭')").all()
+        for btn in close_btns:
+            try:
+                if btn.is_visible():
+                    btn.click()
+                    break
+            except Exception:
+                pass
+        time.sleep(2)
+
+        if "/apikey" not in page.url:
+            page.goto(AI_STUDIO_URL, timeout=15000)
+            time.sleep(3)
+
+    print(f"\n{'='*60}")
+    print(f"🎉 完成! 成功创建 {len(keys_created)}/{args.count} 个 Gemini API Key")
+    for idx, key in enumerate(keys_created, 1):
+        print(f"  Key {idx}: {key}")
+    print(f"\n存储位置: {DB_PATH}")
+    print(f"JSON 目录: {JSON_DIR}")
+    print(f"{'='*60}")
+
+    input("\n按 Enter 关闭浏览器...")
+    browser.close()
+    pw.stop()
+
+
+if __name__ == "__main__":
+    main()

03_卡木（木）/木根_逆向分析/全网AI自动注册/脚本/tokens/gemini_gemini_project_1_at_google.com_1773573688.json

@@ -0,0 +1,7 @@
+{
+  "provider": "gemini",
+  "email": "gemini_project_1@google.com",
+  "api_key": "AIzaSyA6RHlWHTfUEqMzdfQQQSmugsGcoDIxnAg",
+  "project": "project_1",
+  "registered_at": "2026-03-15T11:21:28.308260+00:00"
+}

03_卡木（木）/木根_逆向分析/全网AI自动注册/脚本/tokens/gemini_gemini_project_1_at_google.com_1773573760.json

@@ -0,0 +1,7 @@
+{
+  "provider": "gemini",
+  "email": "gemini_project_1@google.com",
+  "api_key": "AIzaSyA6RHlWHTfUEqMzdfQQQSmugsGcoDIxnAg",
+  "project": "project_1",
+  "registered_at": "2026-03-15T11:22:40.294576+00:00"
+}

03_卡木（木）/木根_逆向分析/全网AI自动注册/脚本/tokens/gemini_gemini_project_2_at_google.com_1773573702.json

@@ -0,0 +1,7 @@
+{
+  "provider": "gemini",
+  "email": "gemini_project_2@google.com",
+  "api_key": "AIzaSyA6RHlWHTfUEqMzdfQQQSmugsGcoDIxnAg",
+  "project": "project_2",
+  "registered_at": "2026-03-15T11:21:42.299331+00:00"
+}

03_卡木（木）/木根_逆向分析/全网AI自动注册/脚本/tokens/gemini_gemini_project_2_at_google.com_1773573774.json

@@ -0,0 +1,7 @@
+{
+  "provider": "gemini",
+  "email": "gemini_project_2@google.com",
+  "api_key": "AIzaSyA6RHlWHTfUEqMzdfQQQSmugsGcoDIxnAg",
+  "project": "project_2",
+  "registered_at": "2026-03-15T11:22:54.880143+00:00"
+}

03_卡木（木）/木根_逆向分析/全网AI自动注册/脚本/tokens/gemini_gemini_project_3_at_google.com_1773573717.json

@@ -0,0 +1,7 @@
+{
+  "provider": "gemini",
+  "email": "gemini_project_3@google.com",
+  "api_key": "AIzaSyA6RHlWHTfUEqMzdfQQQSmugsGcoDIxnAg",
+  "project": "project_3",
+  "registered_at": "2026-03-15T11:21:57.080669+00:00"
+}

03_卡木（木）/木根_逆向分析/全网AI自动注册/脚本/tokens/gemini_gemini_project_3_at_google.com_1773573789.json

@@ -0,0 +1,7 @@
+{
+  "provider": "gemini",
+  "email": "gemini_project_3@google.com",
+  "api_key": "AIzaSyA6RHlWHTfUEqMzdfQQQSmugsGcoDIxnAg",
+  "project": "project_3",
+  "registered_at": "2026-03-15T11:23:09.014341+00:00"
+}

04_卡火（火）/火炬_全栈消息/全栈开发/SKILL.md

@@ -4,8 +4,8 @@ description: 卡若AI 全栈开发（火炬）— 知己及类似项目经验 +
 triggers: 全栈开发/知己项目/分销/存客宝/RAG/向量化/Next.js/知识库/卡若AI官网/官网开发/全站开发/开发文档/1～10/开发模板/官网全站/v0前端/v0生成/毛玻璃/前端规格/神射手/毛狐狸/前端标准/实施计划/两阶段评审/橙色锁/配色/API调用/使用手册
 owner: 火炬
 group: 火
-version: "2.4"
-updated: "2026-03-12"
+version: "2.5"
+updated: "2026-03-15"
 ---
 
 # 全栈开发（火炬）
@@ -125,6 +125,75 @@ updated: "2026-03-12"
 
 **协同**：Word/文档类清洗用火炬「文档清洗」；部署到 v0、同步 GitHub/Vercel 用金盾「Vercel与v0部署流水线」。
 
+### 1.10 埋点统计（全站强制，2026-03-15 沉淀）
+
+**任何新功能、新页面、新按钮上线时，必须同步接入埋点统计。** 这是全站开发的标准动作，与功能代码同等重要。
+
+#### 为什么强制
+
+没有埋点 = 没有数据 = 无法判断功能是否有效。上线后再补埋点往往遗漏大量初期行为数据。**先埋点、再发布** 是卡若AI 全站的铁律。
+
+#### 标准架构（三层）
+
+```
+前端（小程序/Web）           后端 API                    管理后台
+trackClick(module,       →  POST /api/{平台}/track   →  GET /api/admin/track/stats
+  action, target, extra)     存入 user_tracks 表          按 module/action/时间段聚合
+```
+
+#### 数据模型（user_tracks 表）
+
+| 字段 | 类型 | 说明 |
+|:---|:---|:---|
+| id | string/UUID | 主键 |
+| user_id | string | 用户 ID |
+| action | string | 动作类型：`btn_click` / `page_view` / `tab_click` / `share` / `nav_click` |
+| target | string | 具体目标：按钮名、页面名、分享类型 |
+| extra_data | JSONB | 扩展信息，**必须包含 `module`（所属模块）和 `page`（页面标识）** |
+| created_at | timestamp | 自动时间戳 |
+
+#### 前端埋点工具标准（以小程序为例）
+
+```javascript
+// utils/trackClick.js
+function trackClick(module, action, target, extra) {
+  const userId = app.globalData.userInfo?.id || ''
+  if (!userId) return
+  app.request('/api/miniprogram/track', {
+    method: 'POST',
+    data: {
+      userId, action, target,
+      extraData: Object.assign({ module, page: module }, extra || {})
+    },
+    silent: true
+  }).catch(() => {})
+}
+```
+
+Web 端同理，封装为 `trackClick(module, action, target)` 函数，通过 fetch 静默上报。
+
+#### 埋点接入检查清单
+
+每开发一个页面/功能，按此清单逐项确认：
+
+- [ ] 页面所有**可点击按钮/标签**都已调用 `trackClick`
+- [ ] `module` 参数使用统一命名（home / chapters / read / my / vip / wallet / match / referral / search / settings）
+- [ ] `action` 参数使用标准动词（`btn_click` / `page_view` / `tab_click` / `nav_click` / `share`）
+- [ ] `target` 参数能区分具体按钮（如 `购买VIP`、`充值`、`阅读第3章`）
+- [ ] 后端 track API 已注册路由并能正确存储 `extra_data`
+- [ ] 管理后台「分类标签点击统计」面板能展示该模块的数据
+
+#### 管理后台展示标准
+
+管理后台数据概览页须包含「分类标签点击统计」面板：
+- 支持时间段筛选（今日 / 本周 / 本月 / 全部）
+- 按 module 分组展示，每个模块显示 top N 点击项
+- 自动 30 秒刷新
+
+#### 经验来源
+
+Soul 创业实验项目（2026-03-15）首次实施全站埋点，覆盖小程序 9 个页面 + 管理后台统计面板 + 后端聚合 API。详见 `运营中枢/参考资料/项目经验库_知己与类似项目.md`。
+
 ---
 
 ## 二、项目经验库（知己类，必读）
@@ -190,6 +259,7 @@ scripts/
 | **前端开发/前端标准_神射手与毛狐狸** | 布局/颜色/毛玻璃/组件/特效统一标准，所有项目前端开发参考 |
 | **神射手 开发文档 4、前端** | 神射手项目内前端规范、核心组件代码、截图索引 |
 | **Superpowers与全栈开发对比与优化建议** | `运营中枢/参考资料/Superpowers与全栈开发对比与优化建议.md` — 计划粒度、TDD、两阶段评审、分支收尾等优化方向 |
+| **埋点统计标准（Soul项目沉淀）** | 全站埋点三层架构：前端 trackClick → 后端 track API → 管理后台聚合面板；2026-03-15 Soul 创业实验项目首次实施，见本 Skill 1.10 节 |
 
 ---
 

SKILL_REGISTRY.md

@@ -110,7 +110,7 @@
 | M01g | 快手发布 | 木叶 | **快手发布、发布到快手、快手登录、快手上传、kuaishou发布** | `03_卡木（木）/木叶_视频内容/快手发布/SKILL.md` | 逆向 cp.kuaishou.com API 视频发布 |
 | M01h | 多平台分发 | 木叶 | **多平台分发、一键分发、全平台发布、批量分发、视频分发** | `03_卡木（木）/木叶_视频内容/多平台分发/SKILL.md` | 一键分发到5平台（抖音/B站/视频号/小红书/快手），Cookie统一管理 |
 | M02 | 网站逆向分析 | 木根 | 逆向分析、模拟登录 | `03_卡木（木）/木根_逆向分析/网站逆向分析/SKILL.md` | 网站 API 分析、SDK 生成 |
-| M02a | **全网API自动注册** | 木根 | **API注册、自动注册、批量注册、API Key、注册账号、免费API、API池、key池、自动开号** | `03_卡木（木）/木根_逆向分析/全网API自动注册/SKILL.md` | OpenAI/Cursor/Gemini/Groq 等全网 API 自动注册+Key 池管理 |
+| M02a | **全网AI自动注册** | 木根 | **AI注册、自动注册、批量注册、API Key、注册账号、免费API、API池、key池、自动开号、Gemini注册** | `03_卡木（木）/木根_逆向分析/全网AI自动注册/SKILL.md` | OpenAI/Cursor/Gemini/Groq 等全网 AI API 自动注册+Key 池管理 |
 | M03 | 项目生成 | 木果 | 生成项目、五行模板 | `03_卡木（木）/木果_项目模板/项目生成/SKILL.md` | 按五行模板生成新项目 |
 | M04 | 开发模板 | 木果 | 创建项目、初始化模板 | `03_卡木（木）/木果_项目模板/开发模板/SKILL.md` | 前后端项目模板库 |
 | M05 | 个人档案生成器 | 木果 | 个人档案、档案生成 | `03_卡木（木）/木果_项目模板/个人档案生成器/SKILL.md` | 自动生成个人介绍档案 |

运营中枢/工作台/gitea_push_log.md

@@ -360,3 +360,4 @@
 | 2026-03-15 12:31:12 | 🔄 卡若AI 同步 2026-03-15 12:31 | 更新：金仓、运营中枢工作台 | 排除 >20MB: 11 个 |
 | 2026-03-15 15:48:32 | 🔄 卡若AI 同步 2026-03-15 15:48 | 更新：运营中枢工作台 | 排除 >20MB: 11 个 |
 | 2026-03-15 16:00:52 | 🔄 卡若AI 同步 2026-03-15 16:00 | 更新：运营中枢工作台 | 排除 >20MB: 11 个 |
+| 2026-03-15 19:00:56 | 🔄 卡若AI 同步 2026-03-15 19:00 | 更新：卡木、总索引与入口、运营中枢工作台 | 排除 >20MB: 11 个 |

运营中枢/工作台/代码管理.md

@@ -363,3 +363,4 @@
 | 2026-03-15 12:31:12 | 成功 | 成功 | 🔄 卡若AI 同步 2026-03-15 12:31 | 更新：金仓、运营中枢工作台 | 排除 >20MB: 11 个 | [仓库](http://open.quwanzhi.com:3000/fnvtk/karuo-ai) [百科](http://open.quwanzhi.com:3000/fnvtk/karuo-ai/wiki) |
 | 2026-03-15 15:48:32 | 成功 | 成功 | 🔄 卡若AI 同步 2026-03-15 15:48 | 更新：运营中枢工作台 | 排除 >20MB: 11 个 | [仓库](http://open.quwanzhi.com:3000/fnvtk/karuo-ai) [百科](http://open.quwanzhi.com:3000/fnvtk/karuo-ai/wiki) |
 | 2026-03-15 16:00:52 | 成功 | 成功 | 🔄 卡若AI 同步 2026-03-15 16:00 | 更新：运营中枢工作台 | 排除 >20MB: 11 个 | [仓库](http://open.quwanzhi.com:3000/fnvtk/karuo-ai) [百科](http://open.quwanzhi.com:3000/fnvtk/karuo-ai/wiki) |
+| 2026-03-15 19:00:56 | 成功 | 成功 | 🔄 卡若AI 同步 2026-03-15 19:00 | 更新：卡木、总索引与入口、运营中枢工作台 | 排除 >20MB: 11 个 | [仓库](http://open.quwanzhi.com:3000/fnvtk/karuo-ai) [百科](http://open.quwanzhi.com:3000/fnvtk/karuo-ai/wiki) |