Figma Context MCP
Give your coding agent access to your Figma data. Implement designs in any framework in one-shot.
作者提供的说明(英文原文)
homes.com real-estate for Claude — search, property details, photos, compare, mortgage math
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
HOMES_WS_PORT, HOMES_COMMUNITIES_FILE你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json
{
"mcpServers": {
"homes": {
"command": "npx",
"args": [
"-y",
"homes-mcp"
],
"env": {
"HOMES_WS_PORT": "your-value",
"HOMES_COMMUNITIES_FILE": "your-value"
}
}
}
}在终端中运行这条命令
claude mcp add homes --env HOMES_WS_PORT=your-value --env HOMES_COMMUNITIES_FILE=your-value -- npx -y homes-mcp运行前请先填写空着的值。
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"homes": {
"command": "npx",
"args": [
"-y",
"homes-mcp"
],
"env": {
"HOMES_WS_PORT": "your-value",
"HOMES_COMMUNITIES_FILE": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"homes": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"homes-mcp"
],
"env": {
"HOMES_WS_PORT": "your-value",
"HOMES_COMMUNITIES_FILE": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。
你的选择通过 Cookie 保存在这台设备上,无需账户。
HOMES_WS_PORT envHOMES_COMMUNITIES_FILE env这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月8日
注册名称已通过 GitHub 账户 chrischall 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于–,最新版本(2.2.0)于4天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Node.js。
同一分类中有:Figma Context MCP, Claude Real Video, Figma MCP Server
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Give your coding agent access to your Figma data. Implement designs in any framework in one-shot.
Let any LLM watch a video locally — and search everything it has ever watched.
The Figma MCP server brings Figma design context directly into your AI workflow.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
homes.com real-estate access as an MCP server for Claude — search listings, resolve addresses, fetch property details, price/tax history, market reports, saved homes, photo galleries, and run affordability/mortgage math, all via natural language.
⚠️ homes.com does not publish a public consumer API. This server reads the same server-rendered HTML and Schema.org JSON-LD that homes.com itself ships to your browser, routed through your own signed-in browser tab via the ContextMint Bridge browser extension. Every request acts on behalf of your existing session — your cookies, your TLS, your JS context — exactly as if you'd clicked it in the browser yourself. Treat this as informal use of homes.com. Use at your own discretion.
| Tool | Purpose |
|---|---|
homes_search_properties | Search listings by free-text location (city/ZIP/neighborhood). Slugifies the input into homes.com's URL routing (a blank or unsluggable location is an error, not an empty result) and parses the JSON-LD CollectionPage.mainEntity.itemListElement[]. Filters by property_type, listing_type, sort, and a price_min/price_max band. Returns address, price, beds/baths, sqft, primary photo, listing agent + brokerage; flags truncated/total_estimated past the ~40-listing SSR cap. |
homes_get_by_address | Resolve one US street address to its canonical homes.com property URL + opaque hash. Walks structured typeahead → slug → city/zip search-fallback with whole-token street + unit verification. Returns matched_via; degrades to { resolved: false }. |
homes_resolve_addresses | Bulk homes_get_by_address (up to 100 addresses, input order preserved, per-row outcomes). Prefer for batches ≥ 3. |
homes_get_property | Full record for a property by URL. Parses JSON-LD + DOM-side sections: address, lat/lng, beds/baths, sqft, year built, price, status, agent (name/title/profile — telephone/email only with include_agent_contact: true) + brokerage, highlights, schools, HOA (raw + monthly), lot size (sqft + acres), parking, heating/cooling, MLS id/source, tax, days-on-market, price drops, extracted_features. Optional inline price_history/tax_history. |
homes_get_property_photos | Full photo gallery scraped from tags on the detail page (JSON-LD only carries one image). Returns { url, position, alt? } per photo, filtered to the homes.com CDN. |
homes_bulk_get | Fetch up to 200 properties' structured records in one call (per-row errors, input order preserved). |
homes_compare_properties | Side-by-side comparison of 2–8 properties with an aligned summary table. Per-target errors captured per-row. Concurrent fetches. |
homes_get_nearby_listings | The "Homes for Sale Near This Property" cross-link cards from a detail page (For Sale, optionally Rentals). URL + address only. |
homes_get_history | Combined price + tax history in one fetch: listing_events, ownership_events, lien_events, normalized events_normalized, and tax_records. events_normalized sets price_change_pct only on PriceChange rows; other rows carry homes.com's "List to Sale" figure as list_to_sale_pct (no dom / source_mls keys). |
homes_get_property_history | *Deprecated* — price/ownership/lien timelines only. Prefer homes_get_history. |
homes_get_tax_history | *Deprecated* — year-by-year tax records only. Prefer homes_get_history. |
homes_get_market_report | Median / average / $-per-sqft for a market, derived from the sold search page's JSON-LD. A blank or unsluggable location is an error, not an empty market. |
homes_get_saved_homes | The signed-in user's saved (favorited) homes. Auth-gated. |
homes_get_saved_searches | The signed-in user's saved searches. Auth-gated. |
homes_calculate_affordability | Local affordability calculator — max purchase price from income + DTI + rates. No network. |
homes_calculate_mortgage | Local PITI calculator — principal+interest, taxes, insurance, HOA, PMI. No network. |
homes_estimate_rent_vs_buy | Local rent-vs-buy model — you supply monthly_rent (homes.com has no rental signal to impute it; see below). No network. |
homes_healthcheck | Round-trips /robots.txt through the fetchproxy bridge (the shared registerBridgeHealthcheckTool from @chrischall/mcp-utils/fetchproxy) and returns diagnostics: bridge role/port/version, the extension link state (bridge.session_state, pending_pair_code, extension_connected), a classified error.kind (timeout / bridge_down / session_not_ready / protocol / unknown), and a plain-English hint distinguishing "bridge down" from "extension not connected / pair code pending" from "homes.com-side problem." |
homes_get_session_context / homes_register_session / homes_set_active_session | List / register / switch account *labels* for your own bookkeeping. Label only: requests always use whichever account the fetchproxy browser tab is signed into. |
homes.com does NOT publish rental estimates anywhere on its consumer site — there's no rent_zestimate analogue, no "estimated rent" widget on property detail pages, and no comparable-rentals endpoint. The homes_estimate_rent_vs_buy tool exists for the math, but it requires the caller to pass monthly_rent directly (you can't fetch a rent estimate from homes.com to plug in).
For rental signals, use a sibling MCP: `zillow_get_property` carries rent_zestimate on detail records, and `redfin_get_comparable_rentals` returns rental comps by URL. See issue #28 for the feasibility investigation.
By using this MCP server, you acknowledge and agree to the following:
1. This server accesses your own homes.com session. Every request is dispatched through your own browser tab via the ContextMint Bridge extension — your cookies, your TLS, your session. It does not — and cannot — access anyone else's account.
2. [homes.com's Terms of Use](https://www.homes.com/about/terms-of-use) govern your use of this server, just as they govern your direct use of homes.com. The terms prohibit automated crawling without written permission, and IDX listing data is licensed for personal, non-commercial use only. You are agreeing to those terms every time you invoke a tool in this server.
3. Personal, non-commercial use only. This project is not affiliated with, endorsed by, sponsored by, or in partnership with homes.com or CoStar Group. It is a personal automation tool that reads the same server-rendered HTML and Schema.org JSON-LD homes.com ships to your browser. Do not use it to bulk-extract listings, redistribute IDX data, train AI models, populate a competing real-estate product, or for any commercial purpose.
4. Stability is not guaranteed. This server reads Schema.org JSON-LD ( ) and DOM-side tags from SSR pages whose URL conventions (/ - /, /property/ / /) may change without notice. It may break. It may stop working. That's by design — the surface is not theirs to maintain on our behalf.
5. You accept full responsibility for any consequences of using this server in connection with your homes.com access — rate limiting, account suspension, IP blocks, AWS WAF challenges, or any enforcement action homes.com takes. If homes.com objects to your use, stop using this server.
This section is the maintainer's good-faith summary of the terms — it is not legal advice and does not modify or supersede homes.com's actual ToU.
Add to .mcp.json:
{
"mcpServers": { "homes": { "command": "npx", "args": ["-y", "homes-mcp"] } } }
git clone https://github.com/chrischall/homes-mcp
cd homes-mcp npm install npm run build
{
"mcpServers": { "homes": { "command": "node", "args": ["/path/to/homes-mcp/dist/bundle.js"] } } }
homes-mcp talks to your browser through the ContextMint Bridge extension, which is shared across every fetchproxy-based MCP (zillow-mcp, opentable-mcp, resy-mcp, …) and installed separately — it is not bundled here. Get it from the ContextMint Bridge releases:
chrome://extensions → toggle Developer mode → Load unpacked → pick the unzipped folder.Where it comes from. ContextMint Bridge is the fetchproxy browser extension under its new name, from the same maintainer; fetchproxy's own README (Extension) points to it. Its source is public at nullnet-app/contextmint-bridge: build it yourself, or check a release zip against the .sha256 file published beside it (shasum -a 256 -c contextmint-bridge-chrome- .zip.sha256). There is no Chrome Web Store listing yet.
Open homes.com and sign in. That's all the auth this server needs.
┌────────────────┐ stdio ┌──────────────────┐ WS ┌──────────────────┐ fetch() ┌─────────────┐
│ MCP client │◀────────▶│ dist/bundle.js │◀──────▶│ ContextMint │◀────────────▶│ homes.com │ │ (Claude, etc.) │ │ (Homes MCP) │ :37149 │ Bridge │ (real TLS, │ (your tab) │ └────────────────┘ └──────────────────┘ │ (separate) │ cookies) └─────────────┘
The MCP server runs in Node, but every HTTP call to homes.com is dispatched into your live browser tab through the ContextMint Bridge extension. Each request rides your existing session — TLS fingerprint, cookies, and JS execution context all match the page that's already on screen. No headless browser stand-in, no separate identity, no third-party proxy: just your real browser, acting on its own behalf, with the MCP server picking what to ask for.
homes.com's pages are SSR React with no public JSON API — every tool extracts data from the Schema.org JSON-LD block embedded in each page ( ). Search pages put listings in CollectionPage.mainEntity.itemListElement[]; detail pages emit a RealEstateListing graph node with mainEntity (address, size, geo) and offers.offeredBy[] (listing agent). Photos are scraped from the DOM since the JSON-LD only carries one primary image. The client wraps that into the tool surface so callers never have to parse HTML themselves.
npm test # tsc typecheck + vitest, mocked transport, no network
npm run test:watch npm run test:coverage npm run build # tsc --noEmit + esbuild bundle → dist/bundle.js npm run dev # node dist/bundle.js (after build)
MIT