market-analyst

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

• Third-party content exposure detected (high risk: 0.90). The skill loads and parses reddit-sentiment-*.md reports from /docs/ (explicitly citing Reddit r/[subreddits]) which are user-generated, public Reddit posts/comments and the agent is expected to read and analyze those quotes and fields as part of its workflow, exposing it to untrusted third-party content.