agent-browser

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The documentation examples show embedding plaintext secrets (e.g., "password123" and "proxy_password":"pass") directly in CLI JSON/commands, which requires the LLM to output secret values verbatim.

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

• Third-party content exposure detected (medium risk: 0.30). In the required runtime workflow, the agent-browser open/snapshot/execute functions navigate to a caller-supplied url and then ingest the page’s extracted elements_text/elements and document.body.innerText content (e.g., via the capture/workflow templates), so outsider-authored free text from arbitrary web pages can be read at runtime.