mcp-trademark
USPTO Trademark Watcher MCP server enabling trademark search, watch, and alert management from within Claude or other MCP clients.
README
mcp-trademark
USPTO Trademark Watcher — both an MCP server (for Claude/Cursor/Windsurf) and a standalone CLI (for cron + email alerts).
Why this exists
CompuMark and Corsearch charge $50–$500/mo for trademark watching. Smaller brand owners and indie founders want the same coverage at $19/mo, or a $149 lifetime license they install themselves. We give them:
- MCP tools — search USPTO from inside Claude, watch your marks, review alerts conversationally.
- CLI — same features as a script (cron-friendly).
- Local SQLite — no SaaS dependency, runs offline once cached.
- Similarity scoring that catches look-alikes (typo squatting), sound-alikes (phonetic conflicts), and salient-token overlap.
Pricing
- $19/mo per watched mark (managed)
- $149 lifetime self-host bundle (one-time)
- $499 lifetime + 5 marks managed for 12 months
Install (self-host)
cd C:\openclaw-products\mcp-trademark
python -m venv .venv
.\.venv\Scripts\activate
pip install -e ".[dev]"
CLI usage
tmwatch search "Acme"
tmwatch watch "Acme" --classes 009,042 --threshold 0.75 --email alerts@you.com
tmwatch watches
tmwatch scan
tmwatch alerts --days 30
MCP usage (Claude Desktop)
Add to claude_desktop_config.json:
{
"mcpServers": {
"trademark": {
"command": "python",
"args": ["-m", "src.server"],
"cwd": "C:\\openclaw-products\\mcp-trademark"
}
}
}
Then in Claude:
"Watch the trademark 'Pipely' in classes 009 and 042 and alert me on any filings scoring above 0.7."
Test
pytest tests/ -v
Production roadmap
- [ ] Bulk-feed integration. Replace the on-demand
client.searchwith a nightly download of the USPTO weekly bulk feed at https://bulkdata.uspto.gov/ + a local FTS5 index. This catches NEW filings that haven't been indexed by the search API yet. - [ ] Semantic similarity via Claude (catches "Apple Music" vs "Music for the Apple"-style conflicts that string match misses).
- [ ] Email digest sender (currently records alerts; need an SES/Resend integration to actually mail them).
- [ ] Web dashboard (Next.js) for non-technical buyers.
- [ ] TM-5 international support (EUIPO, WIPO Madrid) for global watches.
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。