chrome-devtools

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill contains examples and commands that require embedding secrets verbatim (cookies, Bearer tokens, passwords passed as CLI arguments or hard-coded into scripts), so an agent would need to handle/output secret values directly.

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

• Third-party content exposure detected (high risk: 0.85). Outsider free text from a runtime-fetched page is ingested into the LLM context via aria-snapshot.js/snapshot.js (they page.goto(args.url) then extract accessible names/textContent into YAML/JSON returned to the agent).