CLI Reference
Complete command reference for the ClawButler CLI (`ap`), including runtime-host discovery and operator workflows.
Installation & Setup#
The ClawButler CLI (command: ap) is a Python tool installed via pip or uv:
# Install with pip
pip install clawbutler-cli
# Or with uv (recommended)
uv pip install clawbutler-cli
# Or run from repo
cd apps/cli && uv run ap --helpAfter installation, verify with:
$ ap --versionclawbutler-cli 0.1.0Authentication#
Before using any command, authenticate with your ClawButler instance:
Log in with email/password or API token. Creates a stored profile.
$ ap auth loginRefresh an expired access token using the stored refresh token.
$ ap auth refreshClear stored session tokens for one or all profiles.
$ ap auth logoutShow the current authenticated user's details.
$ ap auth whoamiThe CLI supports multiple profiles for switching between instances:
Show all stored profiles
$ ap profile listSwitch active profile
$ ap profile use <name>Delete a stored profile
$ ap profile remove <name>Command Tree#
Current top-level commands, grouped to roughly match the Web navigation.
Operate Commands
ap dashboardRead dashboard summaries and top-level operational healthap agentList, show, enable, disable, kill-switch, and configure agentsap approvalList, approve, reject, batch-process approvals and manage rulesap kill-switchEmergency stop and recovery (`trigger` / `recover`)ap activityView real-time agent activity feedap sessionList and inspect agent sessionsap chatStart streaming chat with an agent
Insights Commands
ap auditSearch and export audit trail entriesap costView cost summaries, trends, and breakdowns by agent/model/taskap budgetManage budgets, view advice, adopt/dismiss recommendationsap cronList, create, update, delete, and inspect cron job runsap configView config versions, diff, drift check, and rollback
Productivity Commands
ap memoryView and manage agent memory entriesap toolMonitor tool usage, status, and call historyap reportGenerate, list, export, and schedule reportsap templateBrowse, validate, and deploy verified templatesap runbookList, trigger, inspect, and manage runbook executionsap analyticsView usage analytics and performance metrics
System Commands
ap connectorManage connectors, discovery import, diagnostics, sync, and IM channelsap runtime-hostList paired runtime hosts, discover instances, batch-add connectors, and sync attach-mode sessionsap orgManage organizations, members, and rolesap notificationView and manage notification preferencesap health-checkRun system health checks and diagnosticsap infrastructureView infrastructure status and service healthap supply-chainView supply chain security and dependency intel
Utility Commands
ap authLogin, logout, refresh tokens, and check identityap profileManage stored connection profilesap exportExport data (agents, config, audit logs) as JSON/CSVap importImport configuration from fileap routingView and manage A2A routing rulesap capabilitiesDiscover agent capabilities and protocol support
Common Options#
These flags are available on most commands:
- ---json — Output as JSON instead of formatted tables
- ---page — Page number (default: 1)
- ---page-size — Items per page (default: 20, max: 100)
- ---date-from — Start date filter (YYYY-MM-DD)
- ---date-to — End date filter (YYYY-MM-DD)
- ---profile — Use a specific stored profile
- ---confirm — Required for destructive operations (delete, logout)
Example Workflows#
Daily Monitoring
Quickly check the operator state with canonical dashboard, agent, and approval commands:
$ ap dashboard summary
ap agent list --health-status error
ap approval list --status pendingCost Investigation
Drill into cost changes with summary, by-agent, and trend commands:
$ ap cost summary --date-from 2026-04-01
ap cost by-agent --date-from 2026-04-01
ap cost trend --granularity week --date-from 2026-04-01Emergency Response
Inspect one agent, trigger the kill switch if needed, then inspect audit history:
$ ap agent show <agent-id>
ap kill-switch trigger <agent-id> --reason "Unexpected outbound calls"
ap audit search --agent-id <agent-id> --date-from 2026-04-15Runtime Host intake
Complete the paired-host discovery and session-sync loop from terminal only:
$ ap runtime-host list
ap runtime-host discover <runtime-host-id>
ap runtime-host batch-add <runtime-host-id> --instance-key <instance-key>
ap runtime-host sync-sessions <runtime-host-id> --instance-key <instance-key>