flow-next-resolve-pr

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches user-generated GitHub PR review threads and top-level comments via scripts/get-pr-comments and scripts/get-thread-for-comment and then passes those comment bodies and synthesized cluster_brief verbatim into resolver agents (see cluster-analysis.md and the Phase 5 dispatch in workflow.md), so untrusted third‑party content is read and can materially influence automated decisions and actions.