llamaIndex - 💡(How to fix) Fix [Feature Request]: Add ProxyClaw as an Agent Tool / Web Access Integration [1 participants]

Feature Description

Please add ProxyClaw as an official LlamaIndex Agent Tool integration.

What is ProxyClaw? ProxyClaw is a residential proxy + web access layer built specifically for AI agents. It lets AI agents browse the web through 2M+ residential IPs across 195+ countries, bypass CAPTCHAs, Cloudflare, and anti-bot systems, and returns clean Markdown/JSON instead of raw HTML — making it natively LLM-ready.

Why it fits LlamaIndex:

• Agent Tool: enables agents to browse any website, scrape data, and retrieve structured content
• • Returns Markdown/JSON directly — no parsing needed, perfect for RAG pipelines
• • • Anti-bot bypass built in — no config needed (66/66 sites, 100% success rate)
• • • • Free tier (0.5 GB, no credit card) + earn credits by running a Docker node
• • • • • Pricing: $1.50/GB vs competitors at $8-15/GB
• • • • • • Already integrated with LangChain (pip install langchain-proxyclaw) Links:
• Website: https://proxyclaw.ai
• • GitHub: https://github.com/Iploop/proxyclaw
• • • PyPI: https://pypi.org/project/iploop-sdk/
• • • • Docs: https://docs.iploop.io
• • • • • Install: pip install iploop-sdk

Reason

LlamaIndex agents currently lack a native, agent-optimized web browsing tool. Existing solutions (BrightData, Oxylabs) are built for human scrapers — they return raw HTML and require complex parsing. ProxyClaw was designed from the ground up for AI agents, returning clean Markdown/JSON and handling anti-bot systems automatically.

Value of Feature

ProxyClaw gives LlamaIndex agents real-world web access with zero setup — agents can browse, scrape, and retrieve live data from any website without being blocked. With 2M+ residential IPs, 100% anti-bot success rate, and LLM-ready output, it directly enables powerful RAG and agentic workflows that currently aren't possible without complex proxy infrastructure.