tidy

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

• Third-party content exposure detected (high risk: 1.00). The skill's workflow (SKILL.md step 2, "Research unknown transactions") explicitly instructs the agent to perform web searches on merchant names/domains/raw descriptions and use those external results to identify merchants and decide categorization/rule actions, exposing it to untrusted public web content that can influence behavior.