Connect ProofPrice to your AI

Claude (claude.ai and the desktop app)

1. Open Customize > Connectors and click Add custom connector.
2. Paste the server URL: https://proofprice.com/api/mcp
3. Leave sign-in off (No sign-in). ProofPrice needs no key. Click Add.
4. In a chat, click + > Connectors and switch ProofPrice on.
5. On Team and Enterprise plans an Owner adds it under Organization settings > Connectors; members then click Connect.

Official guide: https://www.claude.com/docs/connectors/custom/remote-mcp

ChatGPT (developer mode / apps)

1. Developer mode is on the web for Pro, Plus, Business, Enterprise and Education accounts: Settings > Apps > Advanced settings > turn on Developer mode.
2. Click Create app and enter the remote MCP server URL (Streamable HTTP):
3. https://proofprice.com/api/mcp
4. Choose No Authentication, then create it. It appears under Drafts.
5. In a chat open the + menu, choose Developer mode and pick ProofPrice. Name the tool in your prompt if ChatGPT does not call it.

OpenAI's menus change often (its Apps SDK guide calls this Add custom MCP server under Plugins). If yours differs, follow the official guides below. ProofPrice's tools are read-only.

Official guide: https://developers.openai.com/api/docs/guides/developer-mode

Cursor

1. Add this to .cursor/mcp.json in your project, or ~/.cursor/mcp.json for all projects.

.cursor/mcp.json

{
  "mcpServers": {
    "proofprice": {
      "url": "https://proofprice.com/api/mcp"
    }
  }
}

Official guide: https://cursor.com/docs/mcp

VS Code / GitHub Copilot

1. Add this to .vscode/mcp.json in your workspace, or run the command MCP: Open User Configuration to add it for all workspaces.
2. You can also run MCP: Add Server from the Command Palette and choose HTTP.

mcp.json

{
  "servers": {
    "proofprice": {
      "type": "http",
      "url": "https://proofprice.com/api/mcp"
    }
  }
}

Official guide: https://code.visualstudio.com/docs/agent-customization/mcp-servers

Windsurf

1. In Cascade, use Open MCP config file (the file's location depends on your Windsurf version; the official guide lists it) and add:

mcp_config.json

{
  "mcpServers": {
    "proofprice": {
      "serverUrl": "https://proofprice.com/api/mcp"
    }
  }
}

Official guide: https://docs.windsurf.com/plugins/cascade/mcp

Claude Code

terminal

claude mcp add --transport http proofprice https://proofprice.com/api/mcp

Official guide: https://code.claude.com/docs/en/mcp

OpenAI Responses API (hosted MCP tool)

1. OpenAI's servers call ProofPrice for you. Put the MCP tool in the tools list:

python

from openai import OpenAI

client = OpenAI()
resp = client.responses.create(
    model="YOUR_MODEL",
    tools=[{
        "type": "mcp",
        "server_label": "proofprice",
        "server_url": "https://proofprice.com/api/mcp",
        "require_approval": "never",
    }],
    input="What did terraced houses in Bristol actually sell for recently?",
)
print(resp.output_text)

Official guide: https://developers.openai.com/api/docs/guides/tools-remote-mcp

OpenAI Agents SDK

1. Hosted MCP tool (OpenAI calls ProofPrice):

python

from agents import Agent, HostedMCPTool, Runner

agent = Agent(
    name="Price checker",
    tools=[HostedMCPTool(tool_config={
        "type": "mcp",
        "server_label": "proofprice",
        "server_url": "https://proofprice.com/api/mcp",
        "require_approval": "never",
    })],
)
result = await Runner.run(agent, "What is diesel costing near Milan right now?")
print(result.final_output)

# Or connect from your own process over Streamable HTTP:
# from agents.mcp import MCPServerStreamableHttp
# async with MCPServerStreamableHttp(params={"url": "https://proofprice.com/api/mcp"}) as server:
#     agent = Agent(name="Price checker", mcp_servers=[server])

Official guide: https://openai.github.io/openai-agents-python/mcp/

Vercel AI SDK

1. Install the MCP package: npm install @ai-sdk/mcp

typescript

import { createMCPClient } from "@ai-sdk/mcp";

const mcpClient = await createMCPClient({
  transport: { type: "http", url: "https://proofprice.com/api/mcp" },
});
try {
  const tools = await mcpClient.tools();
  // pass `tools` to generateText / streamText
} finally {
  await mcpClient.close();
}

Official guide: https://ai-sdk.dev/docs/ai-sdk-core/mcp-tools

LangChain

1. Current LangChain (1.4 or newer): pip install "langchain[mcp]"
2. Older projects can keep using the langchain-mcp-adapters package (now archived); its snippet is below too.

python

# langchain[mcp] (1.4+)
from langchain.mcp import MCPAdapter

async with MCPAdapter("https://proofprice.com/api/mcp") as adapter:
    tools = await adapter.list_tools()

# langchain-mcp-adapters (pip install langchain-mcp-adapters)
from langchain_mcp_adapters.client import MultiServerMCPClient

client = MultiServerMCPClient({
    "proofprice": {"transport": "http", "url": "https://proofprice.com/api/mcp"},
})
tools = await client.get_tools()

Official guide: https://docs.langchain.com/oss/python/langchain/mcp

LlamaIndex

1. pip install llama-index-tools-mcp

python

from llama_index.tools.mcp import BasicMCPClient, McpToolSpec

mcp_client = BasicMCPClient("https://proofprice.com/api/mcp")
tools = await McpToolSpec(client=mcp_client).to_tool_list_async()

Official guide: https://developers.llamaindex.ai/python/framework/module_guides/mcp/llamaindex_mcp/

Plain REST (any language)

1. No key, JSON back, browsers may call it (CORS is open). The example below is a real one from the coverage list:

terminal

curl "https://proofprice.com/api/v1/fuel?country=ES&area=42&fuel=diesel"

More requests: https://proofprice.com/api/v1/coverage lists the sources and working examples. Machine-readable: https://proofprice.com/openapi.json and https://proofprice.com/llms-full.txt.

Official guide: https://proofprice.com/openapi.json