setup-agent-routing

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

• Third-party content exposure detected (medium risk: 0.30). In SKILL.md the runtime workflow reads GitHub Projects/issues via gh label list/gh project list and the later referenced skill docs use gh issue view --comments and gh issue list --label ..., meaning outsider-authored issue/comment free text can be ingested without first selecting a specific item.