Wraps Novada's 2M+ residential proxy network in an MCP interface with 10 tools for bypassing anti-bot systems and scraping restricted content. Exposes operations like novada_proxy_fetch for single requests, novada_proxy_batch_fetch for concurrent fetching, novada_proxy_extract for structured data extraction with schema support, and novada_proxy_render for JavaScript-heavy pages via real Chromium. Includes sticky sessions to maintain the same IP across multi-step workflows, city-level geo-targeting across 195 countries, and novada_proxy_search for structured Google results. Reach for this when your agent needs to access content behind Cloudflare, scrape at scale without getting blocked, or pull localized data that standard HTTP clients can't touch. Requires Novada credentials from their dashboard, with optional API key for search features.
claude mcp add --transport stdio io.github.novadalabs-proxy4agents-mcp uvx proxy4agents-mcp