From efd14c6892de5c127d02c947a4d055f213b98995 Mon Sep 17 00:00:00 2001 From: skernelx Date: Tue, 17 Mar 2026 23:01:13 +0800 Subject: [PATCH] Add Exa registration support and update README status --- .gitignore | 1 + README.md | 30 +++++- README_EN.md | 30 +++++- exa_browser_solver.py | 244 ++++++++++++++++++++++++++++++++++++++++++ exa_core.py | 19 ++++ mail_provider.py | 47 +++++--- run.py | 30 ++++-- 7 files changed, 375 insertions(+), 26 deletions(-) create mode 100644 exa_browser_solver.py create mode 100644 exa_core.py diff --git a/.gitignore b/.gitignore index b3d98e6..9d748b7 100644 --- a/.gitignore +++ b/.gitignore @@ -11,6 +11,7 @@ ENV/ # 账号信息 accounts.txt firecrawl_accounts.txt +exa_accounts.txt accounts.csv .env .env.local diff --git a/README.md b/README.md index 58b56b9..fbb8033 100644 --- a/README.md +++ b/README.md @@ -4,9 +4,15 @@ 这是当前的多服务注册器与聚合 API 上游工具，聚焦两件事： -- 注册 `Tavily` / `Firecrawl` key +- 注册 `Firecrawl` / `Exa` key - 验证 key 是否真实可用，并把可用 key 提供给统一搜索层 +## 当前状态 + +- `Firecrawl`：可用 +- `Exa`：可用 +- `Tavily`：按当前项目这轮本地测试，官网已关闭邮件注册入口，暂不可用 + 它不是 `MySearch` 的正式产品仓库。 如果你要部署公开可用的统一搜索控制台、MCP、Skill 和 Social / X 接入，请直接使用独立项目： @@ -24,14 +30,14 @@ ## Features -- 多服务启动台：启动时直接选择 Tavily 或 Firecrawl +- 多服务启动台：启动时可选择 Tavily / Firecrawl / Exa - 自动环境准备：自动检查 `venv`、依赖和浏览器 - 邮箱链路统一：支持 Cloudflare Mail API 和 DuckMail - 多域名可选：启动时选择本轮注册用的域名 - 并发注册：支持批量和并发执行 - 后台浏览器模式：默认 headless，必要时可切前台排查 - 真实可用性验证：拿到 key 后马上调用官方接口 -- 自动上传到代理池：上传时会带上服务标记，服务端能自动识别 Tavily / Firecrawl 并写入各自池子 +- 自动上传到代理池：上传时会带上服务标记，服务端能自动识别 Firecrawl / Exa / Tavily 并写入对应池子 - 适合作为统一搜索网关的上游：给 `MySearch-Proxy` 提供稳定的 key 来源 - 跨平台启动：Windows / macOS / Linux 都能直接跑 @@ -62,6 +68,8 @@ cp .env.example .env 编辑 `.env`，填好邮箱配置和可选上传配置。 +如果已经配置了 `SERVER_URL` 和 `SERVER_ADMIN_PASSWORD`，注册成功后可以自动上传。 + ### 3. Run macOS / Linux: @@ -82,6 +90,20 @@ Windows: start_auto.bat ``` +启动后当前可选： + +```text +1. Tavily +2. Firecrawl +3. Exa +``` + +补充说明： + +- `Exa` 注册成功后会保存到 `exa_accounts.txt` +- `Exa` 的格式是 `email,EMAIL_OTP_ONLY,api_key` +- `Tavily` 选项目前仍保留在启动台里，但当前状态暂不可用 + ## 推荐搭配方式 如果你希望把注册出来的 key 接成公开可用的统一搜索服务，推荐直接搭配： @@ -108,7 +130,7 @@ start_auto.bat 这里的 `tavily-key-generator` 只负责： -- 注册 Tavily / Firecrawl key +- 注册 Firecrawl / Exa key - 验证 key 是否真实可用 - 提供可选聚合 API，供 `MySearch` / `MySearch-Proxy` 调用 diff --git a/README_EN.md b/README_EN.md index 0ed59c2..7760361 100644 --- a/README_EN.md +++ b/README_EN.md @@ -5,9 +5,15 @@ This repository is the multi-service registration toolkit and aggregation API upstream for: -- registering `Tavily` / `Firecrawl` keys +- registering `Firecrawl` / `Exa` keys - validating that the keys actually work before feeding a unified search layer +## Current Status + +- `Firecrawl`: available +- `Exa`: available +- `Tavily`: based on the current local project testing, the official email registration entry is closed, so it is temporarily unavailable + It is not the main product repository for `MySearch`. If you want the public-facing unified search console, MCP, Skill, and Social / X integration, use the standalone project: @@ -25,14 +31,14 @@ integration, use the standalone project: ## Features -- Multi-service launcher: choose Tavily or Firecrawl at startup +- Multi-service launcher: choose Tavily / Firecrawl / Exa at startup - Automatic environment bootstrap: checks `venv`, dependencies, and browsers - Unified mail layer: supports Cloudflare Mail API and DuckMail - Multi-domain support: choose the active domain at runtime - Concurrent registration: batch and parallel runs are supported - Background browser mode: headless by default, visible mode when debugging - Real usability verification: validates the key against the official API -- Automatic upload to proxy pools +- Automatic upload to proxy pools with a per-service marker - A clean upstream for unified search gateways such as `MySearch-Proxy` - Cross-platform startup: Windows, macOS, and Linux @@ -63,6 +69,8 @@ cp .env.example .env Edit `.env` and fill in your mail settings and optional upload settings. +If `SERVER_URL` and `SERVER_ADMIN_PASSWORD` are configured, successful registrations can be uploaded automatically. + ### 3. Run macOS / Linux: @@ -83,6 +91,20 @@ Windows: start_auto.bat ``` +Current launcher options: + +```text +1. Tavily +2. Firecrawl +3. Exa +``` + +Notes: + +- `Exa` accounts are saved to `exa_accounts.txt` +- the saved format is `email,EMAIL_OTP_ONLY,api_key` +- the `Tavily` option is still shown in the launcher, but it is currently unavailable based on local testing + ## Recommended Pairing If you want to turn the generated keys into a public, reusable unified search @@ -111,7 +133,7 @@ ships the runnable implementation. This repo only handles: -- Tavily / Firecrawl key registration +- Firecrawl / Exa key registration - live key validation - optional aggregation API serving for `MySearch` / `MySearch-Proxy` diff --git a/exa_browser_solver.py b/exa_browser_solver.py new file mode 100644 index 0000000..47a8819 --- /dev/null +++ b/exa_browser_solver.py @@ -0,0 +1,244 @@ +""" +使用 Camoufox 完成 Exa 注册 +思路：通过邮箱验证码登录，跳过 onboarding，并提取默认 API Key +""" +import json +import os +import re +import threading +import time + +import requests as std_requests +from camoufox.sync_api import Camoufox + +from config import API_KEY_TIMEOUT, EMAIL_CODE_TIMEOUT, REGISTER_HEADLESS +from mail_provider import get_email_code + +_HERE = os.path.dirname(os.path.abspath(__file__)) +_SAVE_FILE = os.path.join(_HERE, "exa_accounts.txt") +_SAVE_LOCK = threading.Lock() +_ACCOUNT_PASSWORD_LABEL = "EMAIL_OTP_ONLY" +_EXA_AUTH_URL = "https://auth.exa.ai/?callbackUrl=https%3A%2F%2Fdashboard.exa.ai%2F" +_EXA_HOME_URL = "https://dashboard.exa.ai/home" + + +def fill_first_input(page, selectors, value): + """填充第一个存在的输入框""" + for selector in selectors: + if page.query_selector(selector): + page.fill(selector, value) + return selector + return None + + +def click_first(page, selectors): + """点击第一个存在的按钮/链接""" + for selector in selectors: + if page.query_selector(selector): + page.click(selector, no_wait_after=True) + return True + return False + + +def extract_api_key(page): + """从页面文本或 HTML 中提取 Exa API Key。""" + patterns = [] + + try: + patterns.extend(re.findall(r"\b[0-9a-f]{8}(?:-[0-9a-f]{4}){3}-[0-9a-f]{12}\b", page.locator("main").inner_text(), re.I)) + except Exception: + pass + + try: + patterns.extend(re.findall(r"\b[0-9a-f]{8}(?:-[0-9a-f]{4}){3}-[0-9a-f]{12}\b", page.content(), re.I)) + except Exception: + pass + + for candidate in patterns: + return candidate + return None + + +def fetch_api_key_via_dashboard_api(page): + """直接调用已登录 dashboard 的 get-api-keys 接口，优先拿完整 key。""" + try: + payload = page.evaluate( + """ + async () => { + const response = await fetch('/api/get-api-keys', { + method: 'GET', + credentials: 'include', + headers: { + 'accept': 'application/json', + }, + }); + + return { + status: response.status, + body: await response.text(), + }; + } + """ + ) + except Exception: + return None + + if int(payload.get("status") or 0) != 200: + return None + + try: + data = json.loads(payload.get("body") or "{}") + except Exception: + return None + + for item in data.get("apiKeys", []): + candidate = (item.get("id") or "").strip() + if re.fullmatch(r"[0-9a-f]{8}(?:-[0-9a-f]{4}){3}-[0-9a-f]{12}", candidate, re.I): + return candidate + return None + + +def ensure_dashboard_ready(page): + """跳过 onboarding，落到可读 API Key 的 dashboard 页面。""" + if "dashboard.exa.ai" not in page.url.lower(): + page.wait_for_url("**/dashboard.exa.ai/**", timeout=30000, wait_until="domcontentloaded") + + if "/onboarding" in page.url.lower(): + click_first(page, ['button:text-is("Skip")']) + time.sleep(1) + click_first( + page, + [ + 'button:text-is("Yes, I don\\\'t want the $10 in credits anyway!")', + 'button:text-is("Yes")', + ], + ) + page.wait_for_url("**/dashboard.exa.ai/**", timeout=30000, wait_until="domcontentloaded") + time.sleep(2) + + if "/home" not in page.url.lower(): + page.goto(_EXA_HOME_URL, wait_until="networkidle", timeout=30000) + time.sleep(2) + + +def wait_for_api_key(page, timeout=20): + """等待主页 API Key 卡片渲染并显示完整 key。""" + start_time = time.time() + while time.time() - start_time < timeout: + ensure_dashboard_ready(page) + api_key = fetch_api_key_via_dashboard_api(page) + if api_key: + return api_key + + click_first(page, ['button:text-is("Show")']) + time.sleep(1) + api_key = extract_api_key(page) + if api_key: + return api_key + time.sleep(1) + return None + + +def save_account(email, api_key): + """并发注册时串行写入 exa_accounts.txt""" + with _SAVE_LOCK: + with open(_SAVE_FILE, "a", encoding="utf-8") as file_obj: + file_obj.write(f"{email},{_ACCOUNT_PASSWORD_LABEL},{api_key}\n") + + +def verify_api_key(api_key, timeout=30): + """真实调用 Exa API，验证新 key 可用""" + try: + response = std_requests.post( + "https://api.exa.ai/search", + json={ + "query": "api key verification", + "numResults": 1, + }, + headers={ + "x-api-key": api_key, + "Content-Type": "application/json", + "Accept": "application/json", + }, + timeout=timeout, + ) + except Exception as exc: + print(f"❌ API Key 调用测试失败: {exc}") + return False + + if response.status_code == 200: + print("✅ API Key 调用测试通过") + return True + + preview = response.text.strip().replace("\n", " ")[:160] + print(f"❌ API Key 调用测试失败: HTTP {response.status_code}") + if preview: + print(f" 响应: {preview}") + return False + + +def register_with_browser(email, password): + """使用浏览器完成 Exa 邮箱验证码注册""" + print(f"🌐 使用浏览器模式注册 Exa: {email}") + + try: + with Camoufox(headless=REGISTER_HEADLESS) as browser: + page = browser.new_page() + + page.goto(_EXA_AUTH_URL, wait_until="networkidle", timeout=30000) + time.sleep(2) + + email_selector = fill_first_input( + page, + ['input[type="email"]', 'input[placeholder="Email"]', 'input[aria-label="Email"]'], + email, + ) + if not email_selector: + print("❌ Exa 登录页未找到邮箱输入框") + return None + + if not click_first(page, ['button:text-is("Continue")']): + print("❌ Exa 登录页未找到 Continue 按钮") + return None + + page.wait_for_selector('input[placeholder*="verification" i], input[aria-label*="verification" i]', timeout=30000) + print("✅ 到达 Exa 邮箱验证码页") + + code = get_email_code(email, timeout=EMAIL_CODE_TIMEOUT, service="exa") + if not code: + return None + + code_selector = fill_first_input( + page, + ['input[placeholder*="verification" i]', 'input[aria-label*="verification" i]'], + code, + ) + if not code_selector: + print("❌ Exa 验证码页未找到输入框") + return None + + if not click_first(page, ['button:text-is("VERIFY CODE")', 'button:text-is("Verify Code")', 'button:text-is("Verify")']): + page.press(code_selector, "Enter") + + page.wait_for_url("**/dashboard.exa.ai/**", timeout=30000, wait_until="domcontentloaded") + print("✅ Exa 登录成功") + + api_key = wait_for_api_key(page, timeout=API_KEY_TIMEOUT) + if not api_key: + print("⚠️ 未找到 Exa API Key") + return None + + print("🧪 验证 API Key 可用性...") + if not verify_api_key(api_key): + return None + + save_account(email, api_key) + + print("🎉 Exa 注册成功") + print(f" 邮箱: {email}") + print(f" 密码: {_ACCOUNT_PASSWORD_LABEL}") + print(f" Key : {api_key}") + return api_key + except Exception as exc: + print(f"❌ Exa 注册失败: {exc}") + return None diff --git a/exa_core.py b/exa_core.py new file mode 100644 index 0000000..0a2b286 --- /dev/null +++ b/exa_core.py @@ -0,0 +1,19 @@ +""" +Exa 注册统一入口 +""" +from mail_provider import create_email +from exa_browser_solver import register_with_browser + + +def register(email, password): + """统一注册入口""" + return register_with_browser(email, password) + + +if __name__ == "__main__": + email, password = create_email(service="exa") + result = register(email, password) + if result: + print(f"✅ 注册成功: {email}") + else: + print(f"❌ 注册失败: {email}") diff --git a/mail_provider.py b/mail_provider.py index 3775f6b..e5987bc 100644 --- a/mail_provider.py +++ b/mail_provider.py @@ -32,7 +32,7 @@ _DUCKMAIL_DOMAIN_PRIORITY = ( _DUCKMAIL_DOMAIN_CACHE = None _DUCKMAIL_MAILBOX_CACHE = {} _SELECTED_DOMAIN = "" -_SUPPORTED_SERVICES = ("tavily", "firecrawl") +_SUPPORTED_SERVICES = ("tavily", "firecrawl", "exa") def rand_str(n=8): @@ -72,7 +72,11 @@ def _normalize_service(service): def _username_prefix(service): service = _normalize_service(service) - return "fc" if service == "firecrawl" else "tavily" + if service == "firecrawl": + return "fc" + if service == "exa": + return "exa" + return "tavily" def create_email(service="tavily"): @@ -104,13 +108,13 @@ def get_verification_link(email, timeout=120): ) -def get_email_code(email, timeout=120): +def get_email_code(email, timeout=120, service="tavily"): """等待邮箱里的 6 位验证码。""" print(f"📨 等待邮箱验证码（最多 {timeout} 秒）...") return _poll_mailbox( email=email, timeout=timeout, - extractor=_extract_email_code, + extractor=lambda message: _extract_email_code(message, service=service), found_message="✅ 收到 6 位验证码", timeout_message="❌ 等待邮箱验证码超时", error_prefix="读取邮箱验证码失败", @@ -175,16 +179,35 @@ def _extract_verification_link(message): return None -def _extract_email_code(message): +def _extract_email_code(message, service="tavily"): + service = _normalize_service(service) subject = (message.get("subject") or "").lower() - if "verify your identity" not in subject and "verify" not in subject and "tavily" not in subject: - return None - + text = message.get("text") or "" content = _message_content(message) - match = re.search(r"\b(\d{6})\b", content) - if not match: - return None - return match.group(1) + combined = f"{subject}\n{content}".lower() + + if service == "exa": + if "exa" not in combined: + return None + if "verification code" not in combined and "sign in" not in combined: + return None + for source in (text, content): + match = re.search( + r"verification code(?:\s+for\s+exa)?(?:\s+is)?[^0-9]*(\d{6})", + source, + re.IGNORECASE, + ) + if match: + return match.group(1) + else: + if "verify your identity" not in subject and "verify" not in subject and "tavily" not in combined: + return None + + for source in (text, content): + match = re.search(r"\b(\d{6})\b", source) + if match: + return match.group(1) + return None def _iter_messages(email): diff --git a/run.py b/run.py index 0e0e0f6..3c818ae 100755 --- a/run.py +++ b/run.py @@ -125,6 +125,7 @@ from config import ( ) from tavily_core import create_email as create_tavily_email, register as register_tavily from firecrawl_core import register as register_firecrawl +from exa_core import register as register_exa from mail_provider import create_email, get_active_domain, get_configured_domains, set_selected_domain # ────────────────────────────────────────────── @@ -199,9 +200,21 @@ def validate_runtime_config(upload, show_provider_summary=True): return True def print_runtime_summary(service="tavily"): - service_name = "Tavily" if service == "tavily" else "Firecrawl" - output_file = "accounts.txt" if service == "tavily" else "firecrawl_accounts.txt" - account_prefix = "tavily-" if service == "tavily" else "fc-" + service_name = { + "tavily": "Tavily", + "firecrawl": "Firecrawl", + "exa": "Exa", + }.get(service, "Tavily") + output_file = { + "tavily": "accounts.txt", + "firecrawl": "firecrawl_accounts.txt", + "exa": "exa_accounts.txt", + }.get(service, "accounts.txt") + account_prefix = { + "tavily": "tavily-", + "firecrawl": "fc-", + "exa": "exa-", + }.get(service, "tavily-") print(f""" ┌──────────────────────────────────────────┐ │ 多服务自动注册启动台 │ @@ -395,8 +408,10 @@ def register_one(index, total, upload, service="tavily"): if service == "tavily": result = register_tavily(email, password) - else: # firecrawl + elif service == "firecrawl": result = register_firecrawl(email, password) + else: + result = register_exa(email, password) if result and result != "SUCCESS_NO_KEY": if upload: @@ -476,14 +491,17 @@ def prompt_service_choice(): print("\n请选择要注册的服务：") print(" 1. Tavily") print(" 2. Firecrawl") + print(" 3. Exa") while True: - print("请输入选项 (1-2，默认 1): ", end="") + print("请输入选项 (1-3，默认 1): ", end="") raw = input().strip() if raw == "" or raw == "1": return "tavily" elif raw == "2": return "firecrawl" + elif raw == "3": + return "exa" else: print("❌ 请输入有效编号") continue @@ -492,7 +510,7 @@ def main(): service = prompt_service_choice() print_runtime_summary(service) - # Firecrawl 不需要 Solver + # 目前只有 Tavily 需要 Solver need_solver = (service == "tavily") if not validate_runtime_config(False, show_provider_summary=True):