Browse categories
Explore
Fiverr Pro
English
$
USD
I will fix claude code grok bot hermes agent ai automation n8n mcp obsidian open claw
Mary Eliz
About this gig
Build, Fix & Deploy Ruflo Claude Flow & Hermes Agent Swarms
Is your Hermes Agent stuck in a PM2 crash loop, failing on WSL2, or unable to connect to Telegram?
You are in the right place.
What I can build & fix :
- Hermes Agent & PM2 Crash Recovery
- Ruflo Claude Flow Swarm Architecture (Queen & Worker)
- Hive Mind Orchestration (SPARC Method & GOAP Planning)
- Self Healing Agent Swarms & Auto Recovery
- MCP Server & Tool Protocol Integration
- n8n Multi Agent Workflow Automation
- Local LLM Setup (Ollama & DeepSeek Privacy Stack)
- WSL2 & Windows Installation Debugging
- LangGraph to Ruflo Workflow Migration
- Telegram, Discord & Messaging Gateway Setup
- Authentication Loop & Credential Fixes
- Full VPS Deployment & Observability
Technologies I work with :
- Hermes Agent
- Ruflo
- Claude Code
- Claude / Anthropic
- OpenClaw
- Obsidian
- MCP Servers
- n8n
- Ollama
- DeepSeek
- LangGraph
- Telegram API
- Grok Bot
- PM2
- WSL2
- Ubuntu VPS / SSH
- Python / Node.js
Services Included :
- System Diagnosis & Error Logging
- Installation & Configuration
- Swarm Architecture Design
- Workflow Migration & Logic Audit
- Local Model Fine Tuning/Integration
- MCP Tool Configuration
- Testing & Performance Optimization
- Debugging & Security Audit
CONTACT ME
Get to know Mary Eliz
Mary Eliz
openclaw
- FromUnited Kingdom
- Member sinceJun 2026
- Avg. response time2 hours
- Last delivery1 month
Languages
English, Spanish, French
I am an AI automation specialist and multi-agent system
developer building self-hosted autonomous AI stacks for
startups, founders, and businesses. I specialize in
OpenClaw, Hermes Agent, n8n workflow automation, Obsidian
vault integration, MCP server setup, Odysseus workspace
deployment, and auto publishing pipelines. I build complete
multi-agent systems that run 24/7 on your VPS with persistent
memory, OpenRouter multi-model routing, and zero cloud
lock-in. Your data stays on your server. I deliver
production-ready AI agent stacks that work from day one.
FAQ
My Hermes Agent or Ruflo setup is already broken. Can you still fix it?
Yes, I regularly rescue broken Hermes Agent and Ruflo Claude Flow installs, including crashed pm2 processes and corrupted MCP configs. Send me your error logs before ordering.
Do I need coding experience to use this?
No coding required on your end. I handle the VPS setup, Ruflo swarm configuration, and Hermes Agent deployment while you focus on using the finished system.
What happens if something breaks after delivery?
You get post-delivery support included in every package to fix any pm2 crashes, Telegram bot disconnects, or MCP errors that appear after handoff.
How fast is turnaround?
Basic setups typically complete in 2-3 days. Full Ruflo swarm deployments with n8n and MCP integration may take 4-5 days depending on complexity.
Does this work on my VPS provider?
Yes, I deploy on Contabo, DigitalOcean, Hetzner, and most Ubuntu Linux VPS providers via SSH, along with WSL2 environments on Windows.
Which AI models and tools do you support?
Ollama, DeepSeek, Claude, and Nous Research Hermes models are all supported, along with LangGraph migration into Ruflo swarm architecture.
Will my existing Obsidian vault or data be safe during setup?
Yes, I take backups before touching any existing configuration and never overwrite your vault, workflows, or credentials without confirming first.
What specific errors can you fix?
WSL2 install failures, npx init errors, pm2 crash loops, auth loops, credential failures, and Ruflo plugin marketplace issues are all covered.
Can you migrate my LangGraph workflows?
Yes, I migrate existing LangGraph agent logic into a Ruflo Claude Flow swarm without losing your original orchestration logic.
