chrome-devtools

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt shows and encourages embedding plaintext credentials into commands and scripts (e.g., node fill.js --selector "#password" --value "secret" and page.type('#password','secret')), which requires the agent to include secret values verbatim and is an insecure pattern.

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

• Third-party content exposure detected (high risk: 0.85). At runtime the skill navigates to an outsider-supplied URL (e.g., node aria-snapshot.js --url ... / navigate.js --url ...), then injects and executes getAriaSnapshotScript() in that page to extract readable YAML/JSON (accessible names/text/links/values) from the fetched third-party DOM into the agent’s LLM context.