investment-results-collector

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

• Third-party content exposure detected (high risk: 0.70). The skill reads and compiles agent outputs (see Step 3: "Compile key findings from all agents") and the example agent metadata explicitly lists tools like "WebSearch" and "WebFetch" in metadata.json, indicating it will ingest content fetched from open/public web sources (untrusted third-party content) and thus can be exposed to indirect prompt injection.