file-upload-methodology

SUSPICIOUS/HIGH-RISK. The content is internally consistent as an offensive file-upload exploitation methodology, but that purpose itself gives an AI agent concrete exploit and post-upload command-execution capability. No malicious installer or credential-harvesting flow is present, yet the skill materially enables unauthorized intrusion, file disclosure, and remote command execution.

这是一个攻击技术说明文档，不是可执行的软件包代码。文档包含可直接用于文件上传绕过、任意文件写入、Webshell 部署和命令执行的示例，因此内容风险较高；但仅凭该静态文本不能认定其自身包含 npm 恶意代码、自动执行行为或数据窃取功能。应将其中示例仅用于授权安全测试，并在实际上传处理逻辑中实施扩展名和 MIME 白名单、随机化存储名、Web 根目录外存储、路径规范化、ZIP 条目校验、符号链接拒绝、禁用上传目录脚本执行及安全 XML/SVG 解析。

This JSON is a high-risk adversarial artifact that provides actionable instructions for deploying and executing a PHP webshell via file-upload bypasses (Apache .htaccess AddType override), multi-layer validation evasion (extension/Content-Type/magic bytes), and multipart filename path traversal into executable web-root directories. While it contains no executable malware code itself, its content is explicitly tailored for remote code execution and should be treated as unsafe for inclusion in any production or supply-chain context.