debt-ops-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly requires re-emitting the helper script's stdout verbatim (and forbids paraphrase), which forces the agent to reproduce any secret values that might appear in that output.

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

• Third-party content exposure detected (high risk: 0.70). SKILL.md instructs the agent to run the bundled scripts/review.py, whose runtime logic reads Markdown files from the repo registry directory (docs/debt/*.md) and emits their frontmatter/body previews to stdout; those files are outsider-authored if they were not authored by the operating user/organization.