What is online_tools-ai?

TongLing Web Standalone is a browser-based console for AI-assisted web penetration testing, meant for security professionals who want to run authorized security assessments without needing a full desktop application. It is a standalone version pulled out of a larger toolkit called TongLing Toolkit, and it now runs on Windows, Linux, and macOS instead of Windows only. The core idea is a terminal you access through your browser, powered by Claude Code together with a tool called HexStrike MCP, which lets the AI plan and run scanning steps on its own. As scans run, the tool shows an interactive graph of what the AI is doing, keeps a log of background tasks, and automatically writes up findings as Markdown reports you can preview and archive. It can also send updates through chat platforms like DingTalk, Telegram, and QQ, and includes support for remote access so you can monitor scans from a phone or tablet. Out of the box it bundles a free edition of HexStrike, vulnerability scanning templates from Nuclei, a proof-of-concept library called Afrog, and a prebuilt search index for known vulnerabilities. Optional pieces include Claude Code itself and tools for fingerprinting and remote networking. Getting started is simplest with Docker: pull the image, run the container, and the console prints an access URL along with a security token needed to log in. There is also a from-source setup for Python 3.10 or newer and Node.js, plus a Docker Compose option for teams that want persistent logs and storage. The README stresses that the access token, API keys, and any remote-access keys should never be committed to a public code repository. This project targets security researchers and penetration testers who already understand scanning workflows and want an AI-guided, visual way to run and document authorized security tests.