agent-browser

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes examples that embed plaintext secrets (e.g., filling a password "password123" and passing "proxy_password"/"proxy_username" in JSON to the CLI), which would require the agent to include secret values verbatim in generated commands/requests.

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

• Third-party content exposure detected (high risk: 0.90). This skill autonomously navigates to arbitrary external URLs and ingests page content (e.g., via open/snapshot/execute returning elements_text and document.body.innerText) and uses those snapshots to decide and perform follow-up interact actions as shown in SKILL.md and templates/capture-workflow.sh/form-automation.sh, exposing the agent to untrusted third‑party content that could carry indirect prompt injections.