dev-test-chrome

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs reading network requests and page storage (e.g., localStorage.getItem('authToken')) and to capture full API requests/responses, which will expose tokens/passwords to the LLM and can require emitting them verbatim.

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow ingests outsider-authored free text from the web page itself via MCP tools like mcp__claude-in-chrome__read_page / mcp__claude-in-chrome__get_page_text (accessibility tree/page text extracted from whatever URL the agent navigates to), which can include prompt-injection content.