find-journalists

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

• Third-party content exposure detected (medium risk: 0.30). The workflow consumes media list search results and journalist profile rows generated by Medialyst (CLI/MCP), which contain untrusted external author names, articles, and bylines.

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

• Direct money access detected (high risk: 0.70). The skill operates a paid asynchronous media-list research workflow via Medialyst credits (CLI/MCP creation and browser handoff), which constitutes a direct financial purchase/service charge operation. The execution layer relies entirely on user-facing instructions ("Ask for explicit approval to create the credit-bearing job") and agent-checked prompts rather than independent service-layer or execution-layer authentication of payment/approval tokens for each transaction, leaving the operation unprotected by an independent qualifying control.