ia-code-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt requires reviewers to include `quoted code` evidence for findings (see Evidence / Output Format) and explicitly lists "secrets exposure" as a security check, which can force the LLM to reproduce secret values verbatim from code when reporting findings.

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md’s required workflow includes reading PR description/issue/task spec and “prior discussions” via gh pr view/gh api .../pulls/{pr}/comments, which ingests outsider-authored free text (PR/issue/comment bodies) into the LLM context for review.