graphql-methodology

该技能与其声明目的基本一致，但声明目的本身就是为AI代理提供进攻性GraphQL利用方法。未见明显凭据窃取或第三方中转，但其能力集包含枚举、绕过、注入、批量滥用和潜在DoS，足以对真实目标造成未授权访问与影响，应判为高风险的进攻性安全技能而非恶意窃密载荷。

This fragment is a JSON configuration for generating/evaluating offensive GraphQL security guidance. It does not contain executable malicious code, obfuscation, or direct system harm, but it explicitly steers outputs toward actionable schema/user enumeration and WAF-bypass techniques (including GET /graphql?query= and required GraphQL markers like __typename/__SCHEMA). Treat it as a high-risk content supply-chain item for misuse rather than as a traditional malware payload.