niche-researcher

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow for “niche research” necessarily involves gathering community information from outsider-authored sources (e.g., public social posts, web pages, or creator content) and then feeding that readable text into the agent’s LLM context for analysis, which creates an indirect prompt-injection risk.