suggest-agent-rules

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly instructs the agent to read GitHub history (PRs/Issues/review comments) using commands like gh pr list/view under "利用できる情報源", which would fetch public, user-generated content from GitHub that the agent is expected to ingest and analyze.