agent-browser

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill contains examples and commands that embed credentials directly into CLI arguments and fill commands (e.g., plaintext passwords, proxy user:pass, headers, cookie values), which would require the LLM to reproduce secret values verbatim in output.

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 can run agent-browser open <url> followed by agent-browser snapshot/snapshot -i, which causes the LLM to ingest attacker-controlled page text/DOM from whatever URL the user provides.