code-review

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly fetches and reads user-generated GitHub PR content (e.g., via the "gh pr view {number}" and "gh pr diff {number}" steps that read the PR title/body/commits/files) and uses that PR text and diff to drive analysis and agent actions, exposing it to untrusted third-party content that could contain indirect prompt injections.