Open to AI assistants
Ask your assistant to read the collection.
Your assistant can search MCLL's published residences, compare budgets or price per square metre, and open the public listing. No MCLL account is required.
Connect your assistant
Paste this message into Codex, Claude Code, or your coding assistant.
Install the MCLL skill with the command npx skills add mcllrealestate/agent-skills, then add the MCP server https://mcllrealestate.com/api/mcp to my configuration. Check that the skill and MCP tools are available.
Manual MCP setup
Endpointhttps://mcllrealestate.com/api/mcp
codex mcp add mcll --url https://mcllrealestate.com/api/mcpWhat it can read
The MCP tools provide access to public listings only. Drafts, leads, owner data, and admin records stay closed.
Search published listings
Filter sale or rent listings by city, area, type, bedrooms, and budget. MCP accepts place names; REST filters use identifiers.
Open the public file
Return price, size, floor, year, features, location, Markdown description, and the public listing URL.
Compare with Code Mode
Run JavaScript in an isolated Cloudflare sandbox with only the MCLL search and get tools in scope.
- Compare these residences by price per square metre and floor.
- If my budget moves from 8 to 10 million baht, what changes?
- Show me river-view residences in Sathorn ready to move in.
Code Mode for comparisons
Search, compare and calculate in one execution.
The execute tool runs the assistant's JavaScript in an isolated Cloudflare Dynamic Worker. It can call only mcll.search and mcll.get on public listings, with no external network, host files or server credentials.
const { results } = await mcll.search({
type: "sale", city: "Bangkok",
});
return results
.filter((r) => !r.priceOnRequest && r.priceThb > 0 && r.areaSqm > 0)
.map((r) => ({
title: r.title,
url: r.url,
pricePerSqm: Math.round(r.priceThb / r.areaSqm),
}));Technical access
Agent skill
Public instructions for finding and comparing MCLL listings with an assistant.
npx skills add mcllrealestate/agent-skillsGitHubREST API
Read-only listings API, documented in OpenAPI 3.1.
Markdown
Detail pages for published listings, developments, neighbourhoods, and articles, in Markdown.
URL.mdAccept: text/markdownDiscovery
The ARD catalogue describes the MCP server, listings API and public skill.
Prefer a person?
However you search, the Firm answers directly.
Enquire