drissionpage-mcp-browser-automation

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt includes examples that require embedding secret values into generated tool calls (e.g., element_type with a password placeholder) and exposes APIs that can return cookie/localStorage values (browser_cookies_get with show_value=True, storage_get), so an LLM might need to handle or output secrets verbatim.

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md describes runtime browser automation tools (e.g., page_navigate, page_snapshot, element_get_text/html/attribute) that can fetch and extract arbitrary web page text into the agent’s context, which is outsider-authored free text from public web content.