openclaw-admin

What This Skill Does

The openclaw-admin skill provides a powerful command-line interface for managing the OpenClaw multi-agent gateway. Designed for system administrators and advanced users, this tool enables granular visibility and control over the gateway's ecosystem. It simplifies the complexity of multi-agent environments by providing direct insight into agent health, routing configurations, token budget constraints, and scheduled tasks. By acting as an administrative layer, the skill allows for rapid troubleshooting and configuration verification without manually parsing multiple JSON files.

Installation

To integrate this utility into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/atiati82/openclaw-admin

Use Cases

Use this skill when you need to audit your agent landscape, troubleshoot routing failures, or verify system load. It is essential for verifying that new agents are properly enabled, ensuring that context window limitations (budgets) are configured correctly for specific LLM providers, and managing the lifecycle of automated background tasks defined in CRONS.json. It is also the primary tool for verifying connectivity with Ollama during local agent deployment.

Example Prompts

1. "OpenClaw, run an audit of all active agents and show me which ones are currently disabled."
2. "Check the gateway diagnostics and list our current routing rules and active cron jobs."
3. "Inspect the context budget settings for our primary models; are any of them hitting their limits?"

Tips & Limitations

• Config Permissions: Ensure that the user account running OpenClaw has proper read access to the config/ directory and the ../openclaw.json master file.
• Read-Only Focus: This skill is primarily for inspection. While it allows for verification of the gateway state, sensitive modifications to configuration files should be performed with caution.
• Ollama Connectivity: If the skill reports 'Ollama not running', ensure the Ollama service is active on the default port before attempting further diagnostics.
• Formatting: The output scripts use formatted strings; if you are running this in a restricted terminal emulator, some character alignments might vary slightly, but the data will remain accurate.