Browse categories
Explore
Fiverr Pro
English
$
USD
I will build custom claude code ai agents and automations
Julien B
About this gig
Do you want Claude Code to do more than answer questions, and actually run repeatable work on your codebase?
I build custom Claude Code setups: specialised subagents, skills, slash commands, hooks and MCP connections to your own tools. Depending on the package, your team gets one reliable agent or a multi-agent pipeline that reviews code, writes documentation, runs checks and reports results.
Everything is tested on a real repository, documented and handed over so your team can extend it without me. Permissions and approval steps are configured so nothing is committed, pushed or deployed without a human confirming.
I use AI-assisted development daily, lead an AI learning community of 2,000 members and teach nearly 5,000 YouTube subscribers. Message me before ordering with your repository, stack and the task you want automated.
Get to know Julien B
Julien B
Founder of MechaPizzAI, AI Automation and Vibe Coding
- FromFrance
- Member sinceMay 2025
- Avg. response time1 hour
- Last delivery12 months
Languages
English, French, Spanish
I turn business ideas and repetitive processes into reliable digital products. With a Master's background in Artificial Intelligence, I build n8n automations, AI agents, websites, SaaS MVPs, local applications and content systems with modern AI-assisted workflows.
I also lead an AI learning community of 2,000 members and teach nearly 5,000 YouTube subscribers how to create practical AI solutions. You get a clear scope, documented work, transparent communication and a solution tailored to your goals, not a generic AI output.
My Portfolio
FAQ
Can I add extra work after ordering?
Yes. Gig extras can be added at checkout, and anything outside the package scope is quoted as a custom offer before I start. Nothing is charged without your approval.
Do I need my own Claude Code subscription?
Yes. The agents run inside your own account, so the subscription and any API usage stay in your name. I explain the expected running cost before we start.
What exactly do I receive?
The agent files, skills, commands and hooks committed to your repository, plus setup documentation. Everything is plain text you can read, edit and extend yourself.
Will the agent change code without me?
Only if you ask for that. By default I configure permissions and approval steps so nothing is committed, pushed or deployed without a human confirming it first.
Can you connect it to our internal tools?
Yes, through an MCP server or a documented API. I only use supported, authorised access, and I never work around a service's own permissions.
Does this work with other AI coding tools?
The setup targets Claude Code. Prompts and documentation transfer to other tools, but I do not guarantee compatibility with a different agent runtime.
Is my code kept private?
Yes. I work only in the repository you give me access to, share nothing publicly, and remove my access at the end of the order on request.
Do you guarantee the agent never makes a mistake?
No. An agent is a tool, not a replacement for review. I build in checks, approval steps and logging so mistakes are caught before they reach production.
How much of the work is AI-generated?
I use AI-assisted development as an accelerator and personally review the architecture, security and final result. If you prefer a fully manual approach, tell me before ordering.
How quickly do you reply?
I usually reply within a few hours on business days. Quick answers on your side for questions and access are what keep a delivery on schedule.
