code-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly instructs including full diffs/commit lists and to quote hunks verbatim in sub-agent prompts and the final report, which would force the LLM to reproduce any secret values present in repository diffs or files.

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

• Third-party content exposure detected (medium risk: 0.30). The workflow reads outsider-authored free text only via Git commit messages that reference external issues/PRs and then fetches those spec/issues using the integrated issue-tracker workflow, so attacker-controlled text from issue/PR bodies can be ingested at runtime through the spec-source identification path.