pi-mom

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

• Third-party content exposure detected (high risk: 0.90). The skill clearly ingests untrusted, user-generated content — Slack messages are append-logged and synced into context.jsonl (log.jsonl → SessionManager) and user-shared attachments are downloaded to the channel attachments/ directory (and SKILL.md and event JSON files under ./data can be written/loaded), all of which are injected into the LLM prompt and can materially influence tool use and subsequent actions.