qc-plan-monitor

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

• Third-party content exposure detected (low risk: 0.10). The required workflow uses F2 to inspect public Douyin work share links (run.cmd:5, SKILL.md:271), which ingests outsider-authored free text into the agent runtime.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (medium risk: 0.30). The skill manages advertising accounts and plans on Ocean Engine / Qianchuan (巨量千川), which involves querying spend/performance, reporting, and mutating plans/materials (with optional dry-run and submission defaults). However, it does not directly execute payments, transfers, or charges of funds itself, nor does it perform direct financial settlements; its operations relate to advertising management and reporting. Therefore, it constitutes an indirect financial/billing scope (managing ad spend reports and plan setups), making the exposure medium/indirect or non-direct. Since it is not a direct financial operation moving or charging money, financial-security exposure is classified as no/medium per the guidelines (service cost or indirect report/campaign management).