Aigist MCP
Publish and manage shareable HTML/Markdown pages with access control and comments via MCP clients.
README
Aigist MCP
Connect your LLM to Aigist — publish and manage shareable HTML/Markdown pages (with access control and Figma-style comments) directly from Claude, Cursor, or any MCP client.
It's a hosted, remote MCP server — you connect by URL, nothing to run or self-host. Sign in with GitHub or Google on first use (OAuth).
- Server URL:
https://ai-gist.com/mcp(Streamable HTTP) - Auth: OAuth (GitHub / Google) — handled by your client on first use
Tools
| Tool | Description |
|---|---|
publish |
Publish an HTML/Markdown page and get a shareable URL |
list_pages |
List your published pages |
get_page |
Retrieve the source of one of your pages |
update_page |
Update one of your pages in place (same URL) |
delete_page |
Delete one of your pages |
Install
Claude Code (CLI)
claude mcp add --transport http aigist https://ai-gist.com/mcp
Claude Desktop — Custom Connector (Pro / Max / Team / Enterprise)
Settings → Connectors → Add custom connector → URL: https://ai-gist.com/mcp → Add.
A browser opens to sign in. Done.
Claude Desktop — double-click bundle (works on any plan, incl. Free)
Custom Connectors are gated to paid plans; the bundle isn't (it's a local
extension). Download aigist.mcpb, double-click it,
and confirm the install in Claude Desktop (it will note the publisher is
unverified — expected for a self-distributed bundle). Sign in on first use.
The bundle wraps
mcp-remoteand ships its dependencies, so it runs on Claude Desktop's built-in Node — no system Node required.
Cursor
Add to your MCP config (~/.cursor/mcp.json or project .cursor/mcp.json):
{
"mcpServers": {
"aigist": { "url": "https://ai-gist.com/mcp" }
}
}
Any other client (generic bridge)
If your client only speaks stdio MCP, use the mcp-remote bridge:
{
"mcpServers": {
"aigist": { "command": "npx", "args": ["-y", "mcp-remote", "https://ai-gist.com/mcp"] }
}
}
Rebuild the bundle
The committed aigist.mcpb is prebuilt. To rebuild (e.g. after bumping the
version or changing the URL in bundle/manifest.json):
bundle/build.sh
MCP Registry
server.json describes this remote server for the
MCP Registry. Before
publishing: set repository.url to this repo, verify the com.ai-gist
namespace (DNS TXT on ai-gist.com), then publish with the mcp-publisher
CLI. Validate against the current schema first — the registry format evolves.
License
MIT — see LICENSE.
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。