v1.7.0 — Proxy4Agents MCP

What's New

New Tool: agentproxy_extract

Extract structured data from any URL — title, price, description, rating, author, date, image, links, headings. Uses meta tags, Open Graph, JSON-LD, and HTML pattern matching. Zero additional API cost.

agentproxy_extract(url="https://amazon.com/dp/...", fields=["title", "price", "rating", "review_count"])

Renamed to bestproxy4agents-mcp

claude mcp add bestproxy4agents-mcp \
  -e NOVADA_PROXY_USER=your_username \
  -e NOVADA_PROXY_PASS=your_password \
  -- npx -y bestproxy4agents-mcp

Old packages (proxy-veil, bestproxy4agents) are deprecated with redirect.

Bug Fixes

• timeout: 0 no longer silently defaults to 60s
• Retry loop now has exponential backoff (500ms)
• HTTP 429 rate-limit errors surface clear message
• Capability-mismatch warnings included in agent response (not just stderr)
• Status output locale-independent formatting

Free Tier

Sign up at novada.com — no credit card. Free access to Scraper API, Web Unblocker, and residential proxies.

---

Install: npx -y bestproxy4agents-mcp
npm: npmjs.com/package/bestproxy4agents-mcp
6 tools: fetch, extract, session, search, render, status
5 providers: Novada, BrightData, Smartproxy, Oxylabs, Generic HTTP

v1.5.0 — Multi-Provider Proxy Architecture

Multi-provider residential proxy MCP server. See proxy-veil on npm.

Supports: Novada, BrightData, Smartproxy, Oxylabs, and any HTTP proxy.