performance-testing-review-ai-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill instructs the model to ingest full code diffs and generate code examples and review comments (potentially quoting lines from the diff) but does not require or implement redaction, so any secrets present in the diffs could be copied verbatim into the model's output.

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow ingests PR-provided free text—notably the pull request description and code diff (e.g., {pr_description} and {code_diff} / diff = reviewer.get_pr_diff()), which are authored by non-operating users—into the LLM prompt for the AI-assisted review step.