safeline

SafeLine, from chaitin, is a self-hosted web application firewall (WAF) positioned in front of web applications as a reverse proxy. According to the README, clients pass through SafeLine before reaching the origin server; the component filters, monitors and blocks malicious HTTP/S traffic heading to the app, and is described as preventing unauthorized data from leaving it. Policies determine which traffic is treated as malicious and which as safe. Attack types the project says it defends against include SQL injection, XSS, code injection, OS command injection, CRLF injection, LDAP injection, XPath injection, RCE, XXE, SSRF, path traversal, backdoor, bruteforce, HTTP flood and abusive bots. The main features listed in the documentation are: Block Web Attacks: filtering of the web attack classes named above. Rate Limiting: throttling traffic that exceeds configured limits, aimed at DoS attempts, bruteforce, traffic surges and similar abuse. Anti-Bot Challenge: human visitors are allowed through while crawlers and bots are blocked. Authentication Challenge: when enabled, visitors must enter a password or are blocked. Dynamic Protection: HTML and JavaScript code from the web server is dynamically encrypted on each visit. The core capabilities also mention a web access control list. The README illustrates each feature with before/after screenshots for legitimate and malicious users. Deployment and configuration are documented externally: an Install Guide and an AddApplication configuration guide on docs.waf.chaitin.com. A note warns that users in mainland China installing the international edition may be unable to connect to cloud services, and points to Chinese-language installation documentation. A live demo is available for evaluation. Ecosystem integrations listed by the repository include an MCP server for AI-assisted management and control, an Ingress-NGINX plugin for protecting Kubernetes ingress traffic, and a Kong Gateway plugin for protecting APIs and services. Community support runs mainly through Discord, with channels for feedback, FAQ and general questions, plus GitHub issues, a website, documentation, an X account and WeChat. The README also presents the project's own operational and evaluation claims, which are the vendor's stated figures rather than independent measurements: a comparison table against ModSecurity Level 1 and Cloudflare Free across detection, false positive and accuracy on a sample set, statements of more than 400,000 installations worldwide, over 1,000,000 websites protected and over 30,000,000,000 HTTP requests handled daily, a 10.0/10 rating from HelloGitHub, listing in awesome-selfhosted, a Product Hunt feature and recognition by Cyber Press. A commercial PRO edition is offered with pricing information and a 7-day trial. Licensing details are in the repository's LICENSE file.