flake

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow uses gh issue view 8375 ... --json body to read the issue body (outsider-authored free text from GitHub issue comments/edits by others) and then gh issue view ... --json body again after editing, so the LLM ingests that external issue text via the GitHub API.