Getting Started
Your guide from first runtime-host pairing or connector setup to daily operations.
Welcome#
ClawButler is the OpenClaw-first governance layer for agent fleets and local runtimes. It can connect to OpenClaw gateways or pair runtime hosts that discover Hermes, Claude Code, Codex, and OpenClaw instances, then bring selected runtimes into one control plane for monitoring, approvals, cost tracking, and audit trails.
This manual mirrors the sidebar navigation of the ClawButler Web console. Each page documents one module — how to configure it, what you'll see, and how to operate it across Web, CLI, and Mobile.
Architecture Overview#
ClawButler either connects to an OpenClaw gateway via a secure WebSocket connector or pairs a runtime host / edge agent. Once connected, it syncs agents, discovered runtime instances, events, cost data, health status, and attach-mode sessions where supported.
Before You Begin#
Cloud Users
Sign up at clawbutler.cc, verify your email, and log in. No browser-side installation is required; you only pair a runtime host or edge agent on the machine you want to govern.
Self-Hosted Users
Follow the Self-Host Guide to deploy ClawButler. Requires Docker, Node.js 22+, and Python 3.12+.
Runtime Host or OpenClaw Gateway
For Cloud, start with a paired Runtime Host / Edge Agent. For self-hosted or advanced environments, you can also connect an OpenClaw gateway manually. Keep the pair code, gateway URL, and operator token handy.
Your First 5 Minutes#
Module Map#
Each module in the sidebar has a dedicated guide below. Click any module to learn how to configure and use it.
Operate — Agent lifecycle and real-time operations
At-a-glance overview of agents, costs, approvals, and live activity.
Discover, configure, and manage agents from a unified registry.
Review and decide on high-risk agent actions with HITL workflows.
Real-time event feed of all agent actions across your fleet.
Inspect individual agent sessions — messages, tool calls, and costs.
Visual map of agent-to-agent relationships and call chains.
Conversational interface to interact with agents directly.
Insights — Analytics, cost tracking, and compliance
Search, filter, and export the immutable audit trail.
Per-agent cost attribution with trend charts and anomaly alerts.
Set cost limits, track spending, and receive overage alerts.
Monitor scheduled jobs — status, next run, and execution history.
Snapshot versioning with semantic diff and rollback preview.
Usage analytics and operational metrics across your fleet.
Productivity — Automation, templates, and reporting
Persistent agent memory entries with search and management.
Monitor tool availability, usage, and permission boundaries.
Generate and export operational reports on demand.
Browse and deploy verified agent configuration templates.
Multi-step automation with pluggable dispatchers.
System — Configuration, integrations, connectors, and operator settings
Manage organizations, teams, and role-based access.
Pair runtime hosts, connect OpenClaw gateways, and manage discovery or session sync from one page.
Configure notification channels and delivery preferences.
Diagnose connectivity, agent health, and system status.
Manage federation peers, edge agents, and deployment settings.
Complete command reference for the ClawButler CLI tool.