popopo-cli

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

• Third-party content exposure detected (high risk: 0.80). The skill instructs the agent to run popopo CLI commands that read room chat messages, live comments, and watch streams (see SKILL.md Workflow and references/commands.md entries like "spaces message", "spaces messages", "spaces watch", "lives comment", "lives comments", "lives watch") and to parse/act on their JSON output, so it clearly ingests untrusted, user-generated third‑party content that can influence subsequent actions.