recall

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

• Third-party content exposure detected (high risk: 0.70). The skill runs entire search/entire explain, which at runtime ingests repository content and command/transcript text; those sources can include outsider-authored free text (e.g., README/transcripts/commit messages/search results authored by others) into the LLM context, creating indirect prompt-injection risk.