grok-cookie-mcp
Cookie-based MCP server for Grok Imagine experiments, enabling image generation and media post creation using your own Grok.com cookies.
README
grok-cookie-mcp
Cookie-based MCP tools for Grok Imagine experiments.
This is an unofficial reverse-engineered local MCP server. It uses your own browser cookies from grok.com; it does not use an xAI API key.
Status
Working:
check_quotagenerate_imagevia/rest/media/post/createcreate_media_postcreate_share_linkupscale_videowrapperpost/getpolling for media posts- Cookie launcher UI for opening an isolated Chrome session
Experimental / not fully working yet:
- App-chat
/responsesrequests can be rejected by Grok anti-bot rules. - Full video generation is still under investigation. Direct
MEDIA_POST_TYPE_VIDEOcreates a post, but may not start the actual video render. - Imagine Agent tools are present but depend on the app-chat path.
Install
uv sync
Cookies
Copy the example and paste your Cookie-Editor export values:
cp cookies.example.json cookies.json
Never commit cookies.json.
Run MCP server
uv run python server.py
Run cookie launcher
uv run python launcher.py
Open:
http://127.0.0.1:8765
Paste Cookie-Editor JSON and launch an isolated Chrome session.
Tools
check_quota— read Grok Imagine quota infogenerate_image— create image media post and return image URLsgenerate_video— experimental video post creation and pollingimage_to_video— experimental, depends on video flowcreate_media_post— low-level media post create wrappercreate_share_link— create media post share linkupscale_video— call video upscale endpointimagine_agent_start/imagine_agent_send/imagine_agent_read— experimental app-chat agent tools
Security
This project is intended for local, authorized use with your own Grok account. Do not commit cookies, tokens, generated session data, or private outputs.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。