blackmount-mcp
Enables AI assistants to access and search local browser history, bookmarks, open tabs, and downloads for personalized context.
README
blackmount-mcp
Your AI assistant is blind to your browsing life. Claude doesn't know what you researched yesterday, what tabs you have open, or what articles you bookmarked. blackmount-mcp fixes that.
Connect your local browser history, bookmarks, open tabs, and downloads directly to Claude, Cursor, or any MCP client. All data stays on your machine — nothing leaves your disk.
Quick Start
pip install blackmount-mcp
Add to claude_desktop_config.json:
{
"mcpServers": {
"blackmount": {
"command": "blackmount-mcp"
}
}
}
Restart Claude Desktop. Done.
What can you ask?
Once connected, try these in Claude:
- "What was that article about React Server Components I read last Tuesday?"
- "How much time did I spend on Reddit vs GitHub this week?"
- "Find all the ML papers I bookmarked"
- "What tabs do I have open right now?"
- "Summarize what I researched yesterday"
- "Find that pricing page I visited for Vercel — I need to compare it"
- "What did I download from arxiv in the last month?"
- "Show me everything I've browsed about LLM fine-tuning"
- "When do I browse the most — mornings or evenings?"
- "Find that GitHub PR tab I have open somewhere"
Tool Catalog
Browser
| Tool | Description | Example Prompt |
|---|---|---|
search_history |
Search browsing history by keyword | "Find that React article I read last week" |
list_history |
Recent browsing history | "What sites did I visit today?" |
find_page |
Natural language page search | "That TypeScript tutorial I read" |
search_bookmarks |
Search bookmarks by title/URL | "Find my saved Python tutorials" |
list_bookmarks |
Browse bookmark folders | "What's in my Work bookmarks folder?" |
get_open_tabs |
List currently open tabs | "What tabs do I have open?" |
find_tab |
Find a specific open tab | "Find that GitHub PR tab" |
list_downloads |
Recent file downloads | "What did I download this week?" |
Analytics
| Tool | Description | Example Prompt |
|---|---|---|
browsing_summary |
Daily browsing summary | "Summarize my browsing today" |
daily_digest |
Daily digest of local browsing | "What did I do online yesterday?" |
productivity_report |
Weekly productivity summary | "How productive was I this week?" |
domain_time |
Time spent per domain | "Where do I spend my time online?" |
distraction_score |
Focus vs distraction ratio | "Am I getting distracted?" |
most_visited |
Top domains by visits | "My most visited sites this month" |
research_topics |
Clustered research topics | "What have I been researching?" |
browsing_patterns |
Peak hours, trends | "When do I browse the most?" |
Claude Desktop Config
{
"mcpServers": {
"blackmount": {
"command": "blackmount-mcp"
}
}
}
Config file locations:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json - Linux:
~/.config/Claude/claude_desktop_config.json
Cursor
One-click install via the badge at the top of this README, or add to ~/.cursor/mcp.json:
{
"mcpServers": {
"blackmount": {
"command": "blackmount-mcp"
}
}
}
Supported Browsers
| Tool family | Browsers |
|---|---|
| History | Chrome, Firefox, Safari, Edge, Brave, Opera, Vivaldi |
| Bookmarks | Chrome, Firefox, Safari |
| Downloads | Chrome, Firefox, Safari |
| Open tabs | Chrome, Safari (macOS via AppleScript; Chrome session-file fallback elsewhere) |
Privacy
- All data is read locally from your machine
- No network requests, no telemetry, no cloud
- Open source (MIT) — audit the code yourself
Companion Apps
- app.blackmount.ai — web dashboard for visual browsing analytics
- Blackmount for iOS — mobile access to your work context
Development
git clone https://github.com/BlackMount-ai/blackmount-mcp.git
cd blackmount-mcp
pip install -e .
pytest tests/ -v
blackmount-mcp
Contributing
Pull requests welcome. Open an issue to discuss larger changes.
Links
- blackmount.ai — Home
- app.blackmount.ai — Web dashboard
- PyPI — Package
- GitHub — Source
License
MIT
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。