pr

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

• Third-party content exposure detected (high risk: 0.80). The skill's required workflow calls GitHub CLI commands like gh pr view --json title,body,labels,number and gh pr diff --stat, which fetch user-generated PR data from GitHub (an external, untrusted third-party source) and the agent is expected to read and act on that content when creating or updating PRs.