Command Center Quick Start
Command Center is the primary administrator experience for configuring, supervising, protecting, and evolving the RapidClaw deployment.
Start with Dashboard and Arcy
Dashboard summarizes the deployment's runtime, subscription, Microsoft Teams, connectors, models, agents, usage, costs, memory, and health. Arcy is the guided administrator and diagnostic assistant. Use Arcy to create or edit agents and skills, configure connectors, draft scheduled tasks, revise prompts or policies, and investigate runtime issues.
Command Center is the everyday admin surface
Use OpenClaw Admin only when lower-level runtime work is genuinely required. Most configuration and operating work should remain in Command Center so RapidClaw can apply its product, identity, permission, approval, and audit boundaries.
Navigation map
| Area | Pages | Use it for |
|---|---|---|
| Operate | Dashboard, Arcy | Health, readiness, guided administration, building, and diagnosis. |
| Create & automate | Agent Roster, Connectors, Skills Catalog, Scheduled Tasks | Build the team, attach governed data and tools, and define recurring work. |
| Review & safety | Approvals, Permission Rules, Agent Instructions, Models, Email | Control what agents may do, review protected actions, select models, and manage mailboxes. |
| Activity & costs | Activity Log, Agent Journal, Usage & Costs | Review operational evidence, agent work, token usage, and projected spending. |
| Help & rollout | Getting Started, Microsoft Teams, Bug Report | Complete setup, validate Teams readiness, use contextual guidance, and report issues. |
| Settings | Backup & Migration, Setup Wizard, OpenClaw Admin | Preserve or restore the workspace, return to deployment controls, or perform advanced runtime administration. |
A practical first-session checklist
- Open Dashboard and resolve any runtime, connector, model, or Teams readiness warnings.
- Open Agent Roster and distinguish RapidClaw-managed agents from custom agents created for your organization.
- Review Models and confirm the intended model assignment for each active agent.
- Review Permission Rules, Agent Instructions, and Approvals before enabling schedules or sensitive tools.
- Open Scheduled Tasks and leave seeded schedules disabled until their scope, tools, output, and approval boundary have been reviewed.
- Confirm Microsoft Teams readiness before expanding access to users.
- Set up Email only for agents that need a mailbox, beginning in test mode.
- Review Usage & Costs and the Azure cost-alert budget before moving from setup into pilot use.
Working with Arcy
- Describe the desired business outcome in ordinary language. Arcy turns short descriptions into reviewable instructions and asks for missing information.
- Use the guided selectors for agents, deployed models, connectors, tools, schedules, data sources, and policies instead of entering internal identifiers.
- Review the proposed scope and safety boundary before approving a protected change.
- Save work as a draft when it is not ready to activate. Creating a draft is different from authorizing live execution.
- Start a new chat when you need clean context for a separate task; ordinary navigation preserves the current Arcy conversation.
Mobile behavior
The full Command Center is designed for desktop administration. On mobile, RapidClaw intentionally exposes only Emergency Stop and Arcy so an administrator can pause governed execution or begin diagnosis without presenting dense configuration screens on a small display.