update-skill

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

• Third-party content exposure detected (high risk: 0.75). In from-chat / from-dump / usage modes, the skill scans and extracts “learnings” from recent conversation or pasted content, which can include outsider-authored free text (e.g., other participants’ chat messages or pasted transcripts) and then feeds those extracted learnings into the agent’s LLM context for proposing edits.