memory-recall

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

• Third-party content exposure detected (high risk: 0.70). The skill’s workflow uses mi cortex recall "<query>", which returns vector-search “hits with source + excerpt”; those retrieved excerpts are runtime content that were not authored by the operating user (outsider sources), so they can be fed into the agent’s LLM context as retrieved text.