create-pull-request

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

• Third-party content exposure detected (medium risk: 0.30). In scripts/analyze_changes.py, the workflow reads attacker-controlled free text from the target repository’s git data at runtime (commit subjects/bodies extracted via git log --oneline and git log --format=%B), then scripts/generate_pr_body.py incorporates that text into the PR body template.