Agent Browser

Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.

Install only if you need broad browser automation and trust the local agent-browser executable in your PATH. Use it only on sites you are authorized to test or automate, avoid proxy-based ban or rate-limit evasion, use test or least-privilege accounts where possible, and protect or delete saved state files, recordings, screenshots, and extracted page data because they may contain credentials, sessions, or private information.

Type: OpenClaw Skill Name: agent-browser-2 Version: 0.1.0 The skill bundle provides the `agent-browser` tool, which grants the AI agent extensive browser automation capabilities as documented in `SKILL.md`. These capabilities include arbitrary JavaScript execution (`agent-browser eval`), local file access (`agent-browser open file://`, `agent-browser upload`), network request interception and modification (`agent-browser network route`), and the ability to load custom browser executables or extensions (`--executable-path`, `--extension`). While the provided documentation and templates are benign, the broad `allowed-tools: Bash(agent-browser:*)` permission in `SKILL.md` combined with these powerful features presents a significant risk for potential misuse, such as data exfiltration or unauthorized execution, if the agent were to receive a malicious prompt.