DataImpulse MCP Proxy
Enables fetching public HTTP(S) pages through DataImpulse residential proxies with country, city, and session targeting, plus an exit IP check tool. Includes best-effort SSRF safeguards and response size limits for safe bounded access.
README
DataImpulse MCP Proxy
Use DataImpulse residential proxy targeting from Claude Code or OpenCode without routing the rest of your system through a proxy. This educational MCP server provides bounded access to public HTTP(S) pages with best-effort SSRF safeguards.
[!WARNING] Never configure
HTTP_PROXYorHTTPS_PROXYglobally. This server creates a DataImpulseProxyAgentfor each MCP request, so other terminals, editors, package managers, and MCP servers remain unaffected.
Quick Start
Prerequisites: Node.js 22.19.0 or later and a DataImpulse account with a username and password.
git clone https://github.com/Gentleman-Programming/dataimpulse-mcp.git
cd dataimpulse-mcp
npm install
The server reads only DI_USER and DI_PASS at runtime. Copy .env.example for local reference if useful, but configure credentials in your MCP client rather than committing a .env file.
Connect Clients
Claude Code
Run this from the checked-out repository, replacing the placeholders and path:
claude mcp add di-proxy --scope user \
--env DI_USER=your_dataimpulse_username \
--env DI_PASS=your_dataimpulse_password \
-- node /absolute/path/to/dataimpulse-mcp/index.js
--scope user makes the server available to your local Claude Code user. Claude Code stores this configuration, including the supplied credentials, in ~/.claude.json; protect that file and never commit or share it.
OpenCode
Add this entry to your OpenCode configuration, replacing the placeholders and absolute path:
{
"mcp": {
"di-proxy": {
"type": "local",
"command": ["node", "/absolute/path/to/dataimpulse-mcp/index.js"],
"environment": {
"DI_USER": "your_dataimpulse_username",
"DI_PASS": "your_dataimpulse_password"
}
}
}
}
Restart OpenCode after saving the configuration so it starts the stdio server with the new environment.
Tool Reference
fetch_page
Fetches a public HTTP(S) page through DataImpulse. By default, it returns cleaned, readable text; set raw to true when the caller needs HTML.
| Parameter | Required | Description |
|---|---|---|
url |
Yes | Public HTTP(S) URL, up to 2,048 characters. URLs with credentials and local/private destinations are rejected. |
country |
No | Two-letter ISO country code, such as US or ES. |
city |
No | City token using letters, numbers, hyphens, or underscores. Requires country. |
session |
No | Stable session token using letters, numbers, hyphens, or underscores. |
raw |
No | Set to true to return HTML instead of cleaned text. |
{
"url": "https://example.com",
"country": "US",
"session": "research-001"
}
{
"url": "https://example.com",
"country": "ES",
"city": "Madrid",
"raw": true
}
check_exit_ip
Checks the public exit IP selected by DataImpulse without fetching a target page. Use it to verify country or session targeting before a workflow.
| Parameter | Required | Description |
|---|---|---|
country |
No | Two-letter ISO country code. |
session |
No | Session token to check a stable route. |
{
"country": "US",
"session": "research-001"
}
Targeting And Errors
Targeting practices
- Use an explicit
countryfor geo-specific content. - Reuse the same
sessionacross multi-step flows that need a consistent route. - Use
cityonly when necessary: DataImpulse charges city targeting at double the normal rate. - Avoid blind retries. Change one variable, observe the result, and stop when the target rejects the request.
Error actions
| Response | Meaning | Action |
|---|---|---|
407 TRAFFIC_EXHAUSTED |
DataImpulse traffic credit is exhausted. | Add traffic credit, then retry. |
407 THREADS_EXHAUSTED |
The account has more than 2,000 active connections. | Reduce concurrent requests, then retry. |
503 NO_RAY |
No proxy IP matches the requested targeting. | Remove city targeting and retain only the country. |
403 |
The destination blocked the request. | Try another country or a fixed session once; respect the site's rules. |
429 |
The destination applied rate limiting or anti-bot controls. | Try one new session or another country once. If it persists, access the target site directly or use another search engine. Do not retry blindly. |
Security Model
- Best-effort SSRF safeguards: every initial URL and redirect target is checked for an HTTP(S) scheme, no credentials, a permitted hostname, and public-only local DNS resolution. Local, private, metadata, credentialed, and mixed-resolution hosts are rejected, with a maximum of 10 redirects.
- Remote proxy resolution:
ProxyAgentsends the hostname to the remote proxy, which performs its own DNS resolution. DNS rebinding or split-horizon DNS can therefore differ from local validation. Do not use this tool with untrusted URLs in high-security environments; if exposing it to untrusted users, prefer a domain allowlist. - Credentials:
DI_USERandDI_PASSare read only at startup and are never logged or returned to MCP clients. - Response bounds: bodies with an advertised
Content-Lengthabove 1,048,576 bytes (1 MiB) are rejected before reading; streaming bodies are cancelled once they exceed that limit. Successful and error output is then limited to 60,000 characters. - Timeout: every request has a 45-second timeout; each per-request proxy agent is destroyed after use.
License
This project is licensed under the MIT License.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.