scrape-review-schema

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly builds and writes a data.js containing extracted values (and copies HTML pages) and then returns browser-submitted feedback verbatim, which requires the agent to include original data values (which may contain secrets) verbatim in generated files/outputs.

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

• Third-party content exposure detected (high risk: 0.75). Yes: the required workflow reads analysis data and saved HTML pages from {spec_path}/{work_path} (outsider-authored page artifacts like meta.json and analyze-page/*.json), then embeds their contents into data.js/review.html which the browser-side JS uses to generate feedback that is sent back to the agent.