Browse categories
Explore
Fiverr Pro
English
$
USD
I will setup openclaw multi agent routing hermes agent odysseus workspace mcp composio
David Adewuyi
About this gig
What happens when your AI agent tries to handle customer support, research, and code execution all at once? It breaks down. Multi-agent routing fixes that
I design and deploy OpenClaw multi-agent architectures with clean routing, isolated workspaces, and a central dashboard so each agent handles its own domain and nothing overlaps.
What I configure:
- OpenClaw multi-agent routing with role-based task assignment
- Paperclip gateway setup for agent coordination and message routing
- Hermes Agent subagent configuration with individual soul.md and memory files
- ClawHub skill installation per agent (research, coding, communication, data)
- Obsidian shared memory vault so agents can read and write to the same knowledge base
- NemoClaw sandbox for safe code execution
- n8n trigger workflows that route tasks to the right agent automatically
- MCP server connections GitHub, Supabase, Google Workspace, Firecrawl, Tavily
- Dashboard configuration for monitoring all agents in one place
- Docker deployment with systemd daemons for 24/7 uptime
Supported deployments: Ubuntu VPS, Mac Mini, DigitalOcean, Hetzner, Hostinger, Contabo.
Message me your setup and goals before ordering.
Get to know David Adewuyi
David Adewuyi
AI Automation Expert
- FromNigeria
- Member sinceFeb 2026
- Last delivery5 months
Languages
English, French, Spanish
I build and fix AI agent automation systems using OpenClaw, Hermes AI Agent, Claude Code, MCP, n8n, Obsidian, and Jarvis AI. I can help with OpenClaw installation, setup, troubleshooting, private Jarvis assistants, multi-agent routing, MCP servers, n8n workflow automation, VPS deployment, and AI agent integrations. Whether you need to install, configure, connect, debug, or automate your AI stack, I create reliable systems tailored to your workflow. AI automation, agent orchestration, Docker, API integration, workflow automation, AI assistant setup.
My Portfolio
FAQ
How is multi-agent routing different from just having one OpenClaw instance?
A: A single instance handles tasks sequentially and gets confused when multiple unrelated requests come in at once. Multi-agent routing assigns each task to a specialized agent so they run in parallel without interfering. The result is faster execution, cleaner outputs, and less hallucination.
Do I need a powerful server to run multiple agents?
A: Two to three agents with cloud LLMs run fine on a $15-20/month VPS. If you want local Ollama models for all agents, 16GB RAM is recommended.
Can the agents share memory and talk to each other?
A: Yes. I configure a shared Obsidian vault as the common memory layer so agents can read each other's outputs, update shared knowledge, and hand off tasks cleanly.
What does Paperclip do exactly?
Paperclip is the gateway layer that receives all incoming messages and routes them to the correct agent based on rules by topic, by channel, by keyword, or by conversation context. It is what makes the routing intelligent.
Can I add more agents later without rebuilding everything?
A: Yes, the architecture I deliver is modular. Adding a new specialized agent is a configuration change, not a rebuild.
What if I already have OpenClaw installed but routing is broken?
A: Send me your current setup details and error logs. Fixing broken routing is available as an add-on. Most issues come from misconfigured Paperclip rules or conflicting soul.md files, which I can diagnose quickly.
