warden-sweep

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 parses third-party GitHub content (e.g., scripts/index_prs.py calls gh pr list and caches PR metadata and diffs with gh pr diff; scripts/find_reviewers.py calls gh api to search users; scripts/organize.py calls gh issue view and reads comments), and those user-generated PR/issue/diff contents are parsed and used to drive deduplication, reviewer selection, labeling, and branching decisions—so untrusted external content is read and can materially influence tool actions.