The Agent That Controls, Captures, Executes
Jebat Agent unifies OpenClaw's control plane and Hermes' capture-first methodology into one powerful AI agent.
Setup in 30 seconds. Deploy local models. Connect channels. Own your AI.
Two Platforms. One Agent.
Jebat Agent combines the best of both OpenClaw and Hermes into a unified, powerful AI agent that does more than just chat.
From OpenClaw
Control Plane Heritage
Provider Routing
Route tasks across multiple LLM backends with automatic fallback chains
Channel Surfaces
Connect Telegram, Discord, WhatsApp, Slack — all through one gateway
Workstation Access
Manage CLI, VS Code, VPS, and remote surfaces from one interface
Health Monitoring
Real-time posture assessment and system health tracking
From Hermes
Capture-First Methodology
Capture-First Workflow
Understand → Plan → Execute → Verify — structured thinking before action
Direct Execution
Low fluff, implementation-first responses that get things done
Multi-Agent Orchestration
Route work across multiple agents with intelligent delegation
Context Awareness
Remember preferences, decisions, and context across every session
What Jebat Agent Can Do
More than just a chatbot — a complete AI operator for your workflow.
30-Second Setup
One command configures your entire workspace with skills, IDE integration, and environment setup.
8 Local LLM Models
Deploy Gemma 4, Qwen2.5 14B, Hermes3, Phi-3, Llama 3.1, Mistral, CodeLlama, and TinyLlama locally.
IDE Integration
Inject JEBAT context into VS Code, Zed, Cursor, Claude Desktop, Gemini CLI, and more.
Channel Setup
Connect Telegram, Discord, WhatsApp, or Slack with guided configuration wizards.
Migration Tool
Automatically migrate from OpenClaw/Hermes. All configs, skills, and workspace converted seamlessly.
Gateway Management
Control your gateway, check agent health, list skills, and deploy to VPS — all from CLI.
From Zero to Agent in 4 Steps
No complex configuration. No manual setup. Just one command.
Install
Run `npx jebat-agent` from any terminal. No installation needed.
Choose Mode
Quick setup for gateway, or full workspace with skills and IDE.
Configure
Connect channels, deploy local models, integrate your IDE.
Start Working
Your agent is ready. Chat, orchestrate, deploy, build.
Every Command You Need
Full reference for the jebat-agent CLI.
npx jebat-agentInteractive setup wizardnpx jebat-agent --quickQuick setup (gateway only)npx jebat-agent --fullFull workspace + skillsnpx jebat-agent --ide vscodeVS Code / Cursor integrationnpx jebat-agent --channel telegramTelegram bot setupnpx jebat-agent --local-model qwen2.5Deploy Qwen2.5 locallynpx jebat-agent --migrateMigrate from OpenClaw/Hermesnpx jebat-gateway startStart gateway servernpx jebat-gateway statusCheck gateway statusnpx jebat-gateway restartRestart gatewaynpx jebat-gateway logsView gateway logsnpx jebat-setup healthCheck agent healthnpx jebat-setup skillsList available skillsnpx jebat-setup testTest agent connectivityHow Jebat Agent Works
A unified architecture that connects everything.
Gateway Layer
Routes requests across 5+ LLM providers with automatic fallback chains and load balancing.
Agent Core
Capture-first execution with multi-agent orchestration, memory, and context awareness.
Channel Layer
Connect Telegram, Discord, WhatsApp, Slack — all through one unified interface.
Ready to Deploy Your Agent?
One command. 30 seconds. Full AI operator ready to go.