firewall-review

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

• Third-party content exposure detected (high risk: 0.70). SKILL.md describes a runtime workflow that reads and analyzes discovered firewall configuration files (including user-supplied comments/descriptions) and then uses their rule text as evidence in the LLM-generated report, so any outsider-authored config content could be ingested as free text into the agent context (indirect prompt injection risk).