What cities does Miklat MCP currently support?
Currently only `jerusalem` is supported, with 198 public shelters. The `list_cities` tool confirms which cities are available.
@danielrosehill
基本信息
MCP to enable AI agents to guide users towards miklatim tziburim (public shelters) in Israel
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"miklat": {
"type": "streamableHttp",
"url": "https://mcp.jlmshelters.com/mcp"
}
}
}`city`, `query`, `limit?`
`city`, `latitude`, `longitude`, `limit?`
`city`
`city`, `id`
`city`
*(none)*
`city`, `shelter_id`, `origin_latitude`, `origin_longitude`, `app?`
`city`, `shelter_type?`, `min_capacity?`, `accessible?`, `limit?`
`city`, `neighborhood`, `limit?`
Miklat MCP is an MCP (Model Context Protocol) server that enables AI agents to find public shelters (miklatim tziburim) in Israel. It currently supports Jerusalem with 198 shelters, providing tools for searching, filtering, and navigating to shelters.
Connect any MCP-compatible client (e.g., Claude Desktop, Claude Code, Cursor) to the streamable HTTP endpoint at https://mcp.jlmshelters.com/mcp. Add the configuration shown in the README to your client's mcpServers block. Use tools like search_shelters, find_nearest_shelters, get_stats, and get_directions_link via the client.
Currently only `jerusalem` is supported, with 198 public shelters. The `list_cities` tool confirms which cities are available.
Submit a pull request to the upstream [Miklat-MCP-Data](https://github.com/danielrosehill/Miklat-MCP-Data) repository with a GeoJSON file following the documented schema. The MCP server periodically pulls data from that repository.
It uses streamable HTTP transport. The endpoint URL is `https://mcp.jlmshelters.com/mcp`. No authentication or API key is required.
No. The README states that data is gathered periodically and no guarantee of accuracy or completeness is offered. Shelters may change between updates. Users should not rely solely on this data for personal safety or emergency preparedness and should verify with official municipal sources.
Clone the repository, run `npm install`, then `npm run dev` to start a local dev server at `http://localhost:8787`. Deploy to Cloudflare Workers with `npm run deploy`.

B2B competitive intelligence, live sales battlecards, 5-axis ICP prospect fit scoring, and tech stack teardowns over MCP and REST with cited public-web evidence.
Open-source MCP to use Blender with any LLM
YC (S26) | Open Computer History | Continuously record your company computer work, map your workflows, help you find work worth automating, and power your agents' context
Painless E2E Automation for Mobile and Web
A curated list of references for MLOps
评论