Integration request: ProxyClaw — residential proxy layer so Toolhouse agents can browse the web without getting blocked

[nickberger1993-ai]

Integration request: ProxyClaw

Toolhouse agents that need to browse the web run into a consistent blocker: Cloudflare, IP bans, CAPTCHAs, and JS-heavy pages that return unusable HTML. ProxyClaw is a residential proxy network built specifically for AI agents that solves this as a drop-in layer.

What ProxyClaw does

• 2M+ residential IPs, 195+ countries
• • Auto CAPTCHA solving (Cloudflare, reCAPTCHA, hCaptcha)
• • • Returns clean Markdown/JSON — not raw HTML — so agents can work with the output directly
• • • • 100% success rate across 66 tested anti-bot sites
• • • • • Free tier: 0.5GB, no credit card required
• • • • • • $1.50/GB (vs BrightData at $8–15/GB)

Install

pip install iploop-sdk

Usage in 3 lines

from iploop import ProxyClaw
pc = ProxyClaw(api_key="YOUR_KEY")
result = pc.get("https://example.com")  # Returns clean Markdown

LangChain integration also available

pip install langchain-proxyclaw

Links

• GitHub: https://github.com/Iploop/proxyclaw
• • Docs + free tier: https://proxyclaw.ai
• • • PyPI: https://pypi.org/project/iploop-sdk/

Proposed integration

Add ProxyClaw as a listed tool/integration in Toolhouse so that agents doing web browsing, scraping, or research tasks can use it to bypass anti-bot systems without extra infrastructure. Would be a natural fit alongside Toolhouse's existing browser/web tools.