participation-warmup-planner

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

• Third-party content exposure detected (high risk: 0.78). High risk: the skill ingests “public community rules and own-account standing via scripts/connectors/discourse.py, hn.py, bluesky.py, fediverse.py” and “public community rules and participation standing come from public surfaces” (plus user-pasted closed-platform rules), so at runtime it may read outsider-authored free text from public web/forum pages and connector-extracted JSON/text into the LLM context.