ohos-chromium-security-review

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill explicitly ingests and audits "provided C/C++ source code" (e.g., "对提供的 C/C++ 源代码进行深度代码审计"), which means it will read and interpret arbitrary user-supplied/untrusted source files (including comments) as part of its workflow, exposing the agent to potential indirect prompt-injection via crafted code or comments.