SteadyFetch

When you need web scraping that won't break on flaky sites or aggressive bot detection, this gives you five core operations: fetch_markdown for LLM-ready content, fetch_url for raw HTML, check_domain for uptime verification, and cache_stats plus clear_cache for managing stored responses. The retry logic and circuit breaker pattern handle intermittent failures automatically, while the anti-bot bypass helps you get through basic protection layers. With over 5,400 tool calls logged and 99% uptime, it's seeing real production use, mostly for the markdown fetching. Reach for this when you're building agents that need to reliably pull web content without babysitting connection failures or manually cleaning HTML for context windows.