MeshOps Control Plane

Operate and secure mesh workflows across gateways, including plugin install/setup, health verification, invite/join bootstrap, and controlled execution tasks...

This skill appears to implement the MeshOps control-plane it claims, but review and operator caution are recommended before installing or enabling it in production: - Metadata mismatch: the marketplace/registry summary claims no required env vars, but metadata.yaml and SKILL.md expect several gating env vars and system binaries. Ask the maintainer to correct the registry metadata or treat the included metadata.yaml as authoritative. - Powerful runtime actions (download+extract, plugin installs, writing to /opt, restarting gateway) are present. They are gated by OPENCLAW_ALLOW_* flags and allowlists, but you should: - Keep OPENCLAW_ALLOW_DEPLOY_SKILL, OPENCLAW_ALLOW_RUN_CMD, and OPENCLAW_ALLOW_HIGH_RISK set to 0 until you audit the skill and its upstream plugin. - Populate OPENCLAW_RUN_CMD_ALLOWLIST narrowly (or leave empty) to avoid arbitrary command execution. - Restrict OPENCLAW_ALLOWED_CALLERS to trusted operator IDs only. - The skill will auto-register capabilities and may automatically claim tasks. If you do not want autonomous cross-gateway execution or capability advertisement, do not load this skill on agents that should remain passive. - Verify sources before permitting installs: setup-ansible-plugin calls 'openclaw plugins install' (github/npm/path). Confirm the plugin repository (https://github.com/likesjx/openclaw-plugin-ansible) and any artifact URLs used by deploy-skill. - Run preflight in a controlled environment first (the included preflight action reports presence of required binaries and current gate settings). Review the collected logs and test deploy flow in an isolated node before enabling on production gateways. If you want to proceed: require maintainer to fix registry metadata, audit the plugin repo and any artifacts, and enforce strict gate/allowlist settings and minimal privileges for the runtime user.

Type: OpenClaw Skill Name: meshops-control-plane Version: 0.1.7 The bundle provides a distributed orchestration framework with high-risk capabilities, specifically remote command execution (run-cmd.sh) and remote skill deployment (deploy-skill.sh). While the implementation includes several security controls—such as environment-based gates (OPENCLAW_ALLOW_HIGH_RISK), caller allowlists, command allowlists, and SHA256 integrity checks—the inherent nature of these tools allows for Remote Code Execution (RCE) and persistence. There is no evidence of intentional malice or data exfiltration, but the powerful administrative surface area warrants a suspicious classification for a senior analyst review.