learn-pr-learnt

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and reads user-generated GitHub content (pull request inline review comments, review summaries, and issue/PR discussion via repos/{owner}/{repo}/pulls/{number}/comments, reviews, and issues/{number}/comments) and uses those comments to derive lessons that change agent behavior, exposing it to untrusted third-party input.