eudic-to-anki

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly exports data from the third‑party Eudic service (see SKILL.md step 3 and modules/export/README.md using python3 scripts/eudic_export.py and references/openapi.md), preserves Eudic's context_line as source_context, and requires the agent to use that source_context when authoring coach JSON and examples (modules/coach/README.md and references/word-coach-json-prompt.md), so arbitrary user/web-origin content from an external service is ingested and can materially influence the agent's decisions and actions.