skill-creator

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

• Third-party content exposure detected (high risk: 0.80). Yes — the SKILL.md Step 1 detection flow and the included references/dynamic-calling.md explicitly require running live web/data fetches (examples: curl/web_search, ddgs/duckduckgo CLI, yfinance price queries, and discord/telegram readers) that ingest open/public or user-generated content which the agent is expected to read and that can influence routing, decisions, and subsequent actions.