Comprehensive skill for installing, configuring, and managing the OpenClaw ecosystem (Gateway, Channels, Models, Automation, Nodes, and Deployment)

OpenClaw CLI wrapper — gateway, channels, models, agents, nodes, browser, memory, security, automation.

Review this before installing if you expect the wrapper to be a strict safety boundary. Only use it with a trusted local openclaw CLI, avoid granting agents broad access to the wrapper, and treat config, gateway, channel, model, agent, message, security, and approvals commands as capable of changing real settings or account state even when OPENCLAW_WRAPPER_ALLOW_RISKY is not set. Avoid bulk shell-environment import and plaintext API keys unless you understand where those secrets may be used.

Type: OpenClaw Skill Name: openclaw-anything Version: 2.0.0 The OpenClaw AgentSkills skill bundle is a wrapper for the OpenClaw CLI, designed to provide a controlled interface to its functionalities. The `SKILL.md` and `references/security-policy.md` explicitly document a security model that categorizes commands as 'low-risk' or 'high-risk'. The `scripts/openclaw.sh` wrapper script strictly enforces this model by requiring the `OPENCLAW_WRAPPER_ALLOW_RISKY=1` environment variable for all high-risk operations (e.g., shell execution, browser automation, cron job modification, secrets application, node control). There is no evidence of intentional malicious behavior, data exfiltration, persistence mechanisms, or prompt injection attempts within the provided files. The wrapper's design is transparently aimed at adding a security layer to a powerful CLI tool, rather than exploiting it.