runbookhermes-aiops-agent

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

• Third-party content exposure detected (high risk: 0.70). The skill's SKILL.md explicitly ingests user-generated incident content from external channels ("Multi-channel intake: Accepts incidents from Web UI, Alertmanager, Feishu, WeCom, API") and shows examples where incident descriptions/metadata (including arbitrary runbook_url and input_text passed to hermes run) are consumed and then drive remediation/rollback decisions, so untrusted third-party content can materially influence agent actions.