
openclaw-control-center
Local-first, security-first control center for OpenClaw agents — visibility dashboard with readonly defaults, token attribution, collaboration tracing, and safe write operations.
- Installs
- 4.4K
- Last 24h
- —
- Trending rank
- —
- Security
- Warn
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/reason-machines/trending-skills --skill openclaw-control-center- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads openclaw-control-center automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
Security audits
Independent scans of the skill's instructions and bundled files.
SKILL.md
The instructions your agent reads when the skill is triggered.
openclaw-control-center
Skill by ara.so — Daily 2026 Skills collection
OpenClaw Control Center transforms OpenClaw from a black box into a local, auditable control center. It provides visibility into agent activity, token spend, task execution chains, cross-session collaboration, memory state, and document sources — with security-first defaults that keep all mutations off by default.
What It Does
- Overview: System health, pending items, risk signals, and operational summary
- Usage: Daily/7d/30d token spend, quota, context pressure, subscription window
- Staff: Who is actively executing vs. queued — not just "has tasks"
- Collaboration: Parent-child session handoffs and verified cross-session messages (e.g.
Main ⇄ Pandas) - Tasks: Task board, approvals, execution chains, run evidence
- Memory: Per-agent memory health, searchability, and source file editing
- Documents: Shared and agent-core documents opened from actual source files
- Settings: Connector wiring status, security risk summary, update status
Installation
git clone https://github.com/TianyiDataScience/openclaw-control-center.git
cd openclaw-control-center
npm install
cp .env.example .env
npm run build
npm test
npm run smoke:ui
npm run dev:ui
Open:
http://127.0.0.1:4310/?section=overview&lang=zhhttp://127.0.0.1:4310/?section=overview&lang=en
Use
npm run dev:uioverUI_MODE=true npm run dev— more stable, especially on Windows shells.
Project Structure
openclaw-control-center/
├── control-center/ # All modifications must stay within this directory
│ ├── src/
│ │ ├── runtime/ # Core runtime, connectors, monitors
│ │ └── ui/ # Frontend UI components
│ ├── .env.example
│ └── package.json
├── docs/
│ └── assets/ # Screenshots and documentation images
├── README.md
└── README.en.md
Critical constraint: Only modify files inside
control-center/. Never modify~/.openclaw/openclaw.json.
Environment Configuration
Copy .env.example to .env and configure:
# Security defaults — do NOT change without understanding implications
READONLY_MODE=true
LOCAL_TOKEN_AUTH_REQUIRED=true
IMPORT_MUTATION_ENABLED=false
IMPORT_MUTATION_DRY_RUN=false
APPROVAL_ACTIONS_ENABLED=false
APPROVAL_ACTIONS_DRY_RUN=true
# Connection
OPENCLAW_GATEWAY_URL=http://127.0.0.1:PORT
OPENCLAW_HOME=~/.openclaw
# UI
PORT=4310
DEFAULT_LANG=zh
Security Flag Meanings
| Flag | Default | Effect |
|---|---|---|
READONLY_MODE | true | All state-changing endpoints disabled |
LOCAL_TOKEN_AUTH_REQUIRED | true | Import/export and write APIs require local token |
IMPORT_MUTATION_ENABLED | false | Import mutations blocked entirely |
IMPORT_MUTATION_DRY_RUN | false | Dry-run mode for imports when enabled |
APPROVAL_ACTIONS_ENABLED | false | Approval actions hard-disabled |
APPROVAL_ACTIONS_DRY_RUN | true | Approval actions run as dry-run when enabled |
Key Commands
# Development
npm run dev:ui # Start UI server (recommended)
npm run dev # One-shot monitor run, no HTTP UI
# Build & Test
npm run build # TypeScript compile
npm test # Run test suite
npm run smoke:ui # Smoke test the UI endpoints
# Lint
npm run lint # ESLint check
npm run lint:fix # Auto-fix lint issues
TypeScript Code Examples
Connecting to the Runtime Monitor
import { createMonitor } from './src/runtime/monitor';
const monitor = createMonitor({
gatewayUrl: process.env.OPENCLAW_GATEWAY_URL ?? 'http://127.0.0.1:4310',
readonlyMode: process.env.READONLY_MODE !== 'false',
localTokenAuthRequired: process.env.LOCAL_TOKEN_AUTH_REQUIRED !== 'false',
});
// Fetch current system overview
const overview = await monitor.getOverview();
console.log(overview.systemStatus); // 'healthy' | 'degraded' | 'critical'
console.log(overview.pendingItems); // number
console.log(overview.activeAgents); // Agent[]
Reading Agent Staff Status
import { StaffConnector } from './src/runtime/connectors/staff';
const staff = new StaffConnector({ gatewayUrl: process.env.OPENCLAW_GATEWAY_URL });
// Get agents actively executing (not just queued)
const activeAgents = await staff.getActiveAgents();
activeAgents.forEach(agent => {
console.log(`${agent.name}: ${agent.status}`);
// 'executing' | 'queued' | 'idle' | 'blocked'
console.log(`Current task: ${agent.currentTask?.title ?? 'none'}`);
console.log(`Last output: ${agent.lastOutput}`);
});
// Get the full staff roster including queue depth
const roster = await staff.getRoster();
Tracing Cross-Session Collaboration
import { CollaborationTracer } from './src/runtime/connectors/collaboration';
const tracer = new CollaborationTracer({ gatewayUrl: process.env.OPENCLAW_GATEWAY_URL });
// Get parent-child session handoffs
const handoffs = await tracer.getSessionHandoffs();
handoffs.forEach(handoff => {
console.log(`${handoff.parentSession} → ${handoff.childSession}`);
console.log(`Delegated task: ${handoff.taskTitle}`);
console.log(`Status: ${handoff.status}`);
});
// Get verified cross-session messages (e.g. Main ⇄ Pandas)
const crossSessionMessages = await tracer.getCrossSessionMessages();
crossSessionMessages.forEach(msg => {
console.log(`${msg.fromAgent} ⇄ ${msg.toAgent}: ${msg.messageType}`);
// messageType: 'sessions_send' | 'inter-session message'
});
Fetching Token Usage and Spend
import { UsageConnector } from './src/runtime/connectors/usage';
const usage = new UsageConnector({ gatewayUrl: process.env.OPENCLAW_GATEWAY_URL });
// Today's usage
const today = await usage.getUsageSummary('today');
console.log(`Tokens used: ${today.tokensUsed}`);
console.log(`Cost: $${today.costUsd.toFixed(4)}`);
console.log(`Context pressure: ${today.contextPressure}`);
// contextPressure: 'low' | 'medium' | 'high' | 'critical'
// Usage trend over 7 days
const trend = await usage.getUsageTrend(7);
trend.forEach(day => {
console.log(`${day.date}: ${day.tokensUsed} tokens, $${day.costUsd.toFixed(4)}`);
});
// Token attribution by task (who ate the scheduled task tokens)
const attribution = await usage.getTokenAttribution();
attribution.tasks.forEach(task => {
console.log(`${task.title}: ${task.tokensUsed} (${task.percentOfTotal}%)`);
});
Reading Memory State
import { MemoryConnector } from './src/runtime/connectors/memory';
const memory = new MemoryConnector({
openclawHome: process.env.OPENCLAW_HOME ?? '~/.openclaw',
});
// Get memory health per active agent (scoped to openclaw.json)
const memoryState = await memory.getMemoryState();
memoryState.agents.forEach(agent => {
console.log(`${agent.name}:`);
console.log(` Available: ${agent.memoryAvailable}`);
console.log(` Searchable: ${agent.memorySearchable}`);
console.log(` Needs review: ${agent.needsReview}`);
});
// Read daily memory for an agent
const dailyMemory = await memory.readDailyMemory('main-agent');
console.log(dailyMemory.content);
// Edit memory (requires READONLY_MODE=false and valid local token)
await memory.writeDailyMemory('main-agent', updatedContent, { token: localToken });
Checking Wiring Status
import { WiringChecker } from './src/runtime/connectors/wiring';
const wiring = new WiringChecker({ gatewayUrl: process.env.OPENCLAW_GATEWAY_URL });
const status = await wiring.getWiringStatus();
status.connectors.forEach(connector => {
console.log(`${connector.name}: ${connector.status}`);
// status: 'connected' | 'partial' | 'disconnected'
if (connector.status !== 'connected') {
console.log(` Fix: ${connector.nextStep}`);
}
});
Approving Tasks (Gated Endpoint)
import { TaskConnector } from './src/runtime/connectors/tasks';
const tasks = new TaskConnector({
gatewayUrl: process.env.OPENCLAW_GATEWAY_URL,
approvalActionsEnabled: process.env.APPROVAL_ACTIONS_ENABLED === 'true',
approvalActionsDryRun: process.env.APPROVAL_ACTIONS_DRY_RUN !== 'false',
});
// This throws if APPROVAL_ACTIONS_ENABLED=false (default)
try {
const result = await tasks.approveTask('task-id-123', { token: localToken });
if (result.dryRun) {
console.log('Dry run — no actual state change');
}
} catch (err) {
if (err.code === 'APPROVAL_ACTIONS_DISABLED') {
console.log('Set APPROVAL_ACTIONS_ENABLED=true to enable approvals');
}
}
UI Section Navigation
Navigate via query params:
http://127.0.0.1:4310/?section=overview&lang=zh
http://127.0.0.1:4310/?section=usage&lang=en
http://127.0.0.1:4310/?section=staff&lang=zh
http://127.0.0.1:4310/?section=collaboration&lang=en
http://127.0.0.1:4310/?section=tasks&lang=zh
http://127.0.0.1:4310/?section=memory&lang=en
http://127.0.0.1:4310/?section=documents&lang=zh
http://127.0.0.1:4310/?section=settings&lang=en
Sections: overview | usage | staff | collaboration | tasks | memory | documents | settings
Languages: zh (Chinese, default) | en (English)
Integration Patterns
Embedding in an Existing OpenClaw Workflow
If your OpenClaw agent needs to hand off instructions to the control center for setup, use the documented install block:
// In your OpenClaw agent task
const installInstructions = `
cd openclaw-control-center
npm install
cp .env.example .env
# Edit .env: set OPENCLAW_GATEWAY_URL and OPENCLAW_HOME
npm run build && npm test && npm run dev:ui
`;
Adding a Custom Connector
All connectors live in control-center/src/runtime/connectors/. Follow this pattern:
// control-center/src/runtime/connectors/my-connector.ts
import { BaseConnector, ConnectorOptions } from './base';
export interface MyData {
id: string;
value: string;
}
export class MyConnector extends BaseConnector {
constructor(options: ConnectorOptions) {
super(options);
}
async getData(): Promise<MyData[]> {
// Always check readonly mode for any write operation
this.assertNotReadonly('getData is readonly-safe');
const response = await this.fetch('/api/my-endpoint');
return response.json() as Promise<MyData[]>;
}
}
Custom UI Section
// control-center/src/ui/sections/MySection.tsx
import React from 'react';
import { useConnector } from '../hooks/useConnector';
import { MyConnector } from '../../runtime/connectors/my-connector';
export const MySection: React.FC = () => {
const { data, loading, error } = useConnector(MyConnector, 'getData');
if (loading) return <div>Loading...</div>;
if (error) return <div>Error: {error.message}</div>;
return (
<ul>
{data?.map(item => (
<li key={item.id}>{item.value}</li>
))}
</ul>
);
};
Troubleshooting
"Missing src/runtime" or "Missing core source"
This almost always means the working directory is wrong:
# Ensure you're in the repo root
pwd # should end in /openclaw-control-center
ls control-center/src/runtime # should exist
If cloned correctly but still missing: the clone was incomplete. Re-clone:
git clone https://github.com/TianyiDataScience/openclaw-control-center.git
UI Doesn't Start / Port Conflicts
# Check if port 4310 is in use
lsof -i :4310
# or on Windows
netstat -ano | findstr :4310
# Change port in .env
PORT=4311
Data Not Showing (Partial or Empty Sections)
- Open
Settings→ 接线状态 (Wiring Status) — it lists exactly which connectors are connected, partial, or missing. - Common causes:
OPENCLAW_GATEWAY_URLnot set or wrong portOPENCLAW_HOMEdoesn't point to actual~/.openclaw- OpenClaw subscription snapshot not at default path
Token Auth Failures
# Generate a local token (see openclaw docs for token location)
cat ~/.openclaw/local-token
# Pass via header in API calls
curl -H "X-Local-Token: <token>" http://127.0.0.1:4310/api/tasks/approve
Approval Actions Silently Do Nothing
Check your .env:
APPROVAL_ACTIONS_ENABLED=true # Must be true
APPROVAL_ACTIONS_DRY_RUN=false # Must be false for real execution
Both must be explicitly set. Default is disabled + dry-run.
Memory Section Shows Inactive Agents
The memory section is scoped to agents listed in openclaw.json. If deleted agents still appear:
# Check active agents
cat ~/.openclaw/openclaw.json | grep -A5 '"agents"'
Remove stale entries from openclaw.json — the memory section will update on next load.
Windows Shell Issues
Prefer npm run dev:ui over UI_MODE=true npm run dev. Cross-env variable setting behaves differently in PowerShell/CMD. The dev:ui script handles this internally.
Prerequisites
- Node.js + npm
- A running OpenClaw installation with accessible Gateway
- Read access to
~/.openclawon the local machine - (Optional)
~/.codexand OpenClaw subscription snapshot for full usage data
FAQ
What does the openclaw-control-center skill do?
Local-first, security-first control center for OpenClaw agents — visibility dashboard with readonly defaults, token attribution, collaboration tracing, and safe write operations.
How do I install openclaw-control-center?
Run npx skills add https://github.com/reason-machines/trending-skills --skill openclaw-control-center in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/openclaw-control-center/ (personal) or .claude/skills/openclaw-control-center/ (project).
Which AI agents can use openclaw-control-center?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is openclaw-control-center safe to install?
Security partners have reviewed it: Gen Agent Trust Hub flagged items to review; Socket passed; Snyk passed. As with any skill, read the SKILL.md and bundled scripts before installing.
Who maintains openclaw-control-center?
It is published by reason-machines in reason-machines/trending-skills. It has 4,419 recorded installs.
More from reason-machines/trending-skills
inkos-multi-agent-novel-writing
reason-machines/trending-skills
Multi-agent CLI system for autonomous novel writing, auditing, and revision with human review gates
claude-hud-statusline
reason-machines/trending-skills
A Claude Code plugin that displays a real-time HUD showing context usage, active tools, running agents, and todo progress in your terminal statusline.
understand-anything-knowledge-graph
reason-machines/trending-skills
Turn any codebase into an interactive knowledge graph using Claude Code skills — explore, search, and ask questions about any project visually.
agency-agents-ai-specialists
reason-machines/trending-skills
A collection of specialized AI agent personalities for Claude Code, Cursor, Aider, Windsurf, and other AI coding tools — covering engineering, design, marketing, sales, and more.
agent-browser-automation
reason-machines/trending-skills
Headless browser automation CLI for AI agents using native Rust binary with Chrome DevTools Protocol
ui-ux-pro-max-skill
reason-machines/trending-skills
AI design intelligence skill for building professional UI/UX across multiple platforms with 161 reasoning rules, 67 styles, and automated design system generation
Popular data & databases skills
supabase-postgres-best-practices
supabase/agent-skills
Postgres best practices maintained by Supabase, for Postgres running anywhere. Load this skill BEFORE writing or changing anything that lives in a Postgres database: creating or altering tables and columns (including choosing column types), schema design, migrations and declarative schema files, RLS policies and the tests that verify them, indexes, triggers, database functions, queues and scheduled jobs (pgcron, pgmq), vector/semantic search (pgvector), and restoring dumps (pgrestore) or importing data. Also load it when diagnosing slow queries, high CPU, timeouts, EXPLAIN plans, connection exhaustion, locking, bloat, or rows visible to the wrong user or tenant. This is not just a performance guide — schema, migration, security, and SQL authoring tasks need these rules too, even for a one-column change or a single query.
prisma-client-api
prisma/skills
Prisma Client API reference covering model queries, filters, operators, and client methods. Use when writing database queries, using CRUD operations, filtering data, or configuring Prisma Client. Triggers on "prisma query", "findMany", "create", "update", "delete", "$transaction".
prisma-database-setup
prisma/skills
Deprecated compatibility name for prisma-orm-setup. Use only when a user or existing integration explicitly invokes prisma-database-setup; otherwise select prisma-orm-setup.
prisma-cli
prisma/skills
Prisma ORM 7 CLI reference for existing Prisma 7 applications or explicitly selected Prisma 7 workflows. Covers prisma init, generate, migrate, db, dev, studio, and mcp. For setup, connection repair, or Prisma 8, use prisma-orm-setup; not the Prisma Platform CLI.
prisma-postgres
prisma/skills
Deprecated compatibility name for prisma-postgres-setup. Use only when a user or existing integration explicitly invokes prisma-postgres; otherwise select prisma-postgres-setup.
prisma-driver-adapter-implementation
prisma/skills
Required reference for Prisma ORM 7 SQL driver adapter work. Use when implementing or modifying adapters, adding database drivers, or touching SqlDriverAdapter, Transaction, savepoint, result mapping, or DriverAdapterError behavior. Covers current transaction lifecycle, optional savepoint hooks, original database-error preservation, and verification.