gh-address-cr

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly ingests and acts on untrusted, user-generated content—e.g., incoming-findings.json / incoming-findings.md produced by external reviewers (normalized in scripts/cli.py::normalize_review_handoff) and GitHub review threads processed/acted on by cr_loop.py and batch_github_execute.py (via the gh CLI)—and that content is read by the agent and can drive replies, resolutions, and fix actions, so third-party inputs can materially influence tool usage and next actions.