review

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

• Third-party content exposure detected (high risk: 0.75). The required workflow runs scripts/diff-analyzer.py, which executes git diff/git diff --cached and parses the resulting unified diff text (including added lines) into JSON that is then used by the agent, so any outsider-authored code/PR content present in the diff can become LLM-readable free text via the diff output.