Browse categories
Explore
Fiverr Pro
English
$
USD
I will build tested claude code agent skills from real research
Chris D
About this gig
Ask an AI to write a skill and it writes what the model already knows. It reads well, and your agent behaves exactly as before.
We research the topic first, then show you the skill working. We have built in the Agent Skills format since it launched in December 2025 and keep a library of 174 skills of our own.
Tell us what your agents should get better at, and give us two or three real questions the skill has to handle. Our researchers read the field (talks, books, reports) alongside anything you send, and write a skill that tells your agent what it does not already know.
You get:
- THE SKILL: an Agent Skills folder, tuned to trigger on your work and stay quiet on the rest. Loads in Claude Code, Codex, Copilot, Cursor and about forty other tools.
- YOUR QUESTIONS, ANSWERED: each one run with the skill and without it, side by side, so you see exactly what it adds.
- VALIDATED AND EVALUATED: every claim checked, and the skill evaluated following the Agent Skills best practices.
Already have skills? Our assess and refresh gig brings them up to date.
Unsure which package? Send your brief first.
Get to know Chris D
Chris D
AI Expert
- FromUnited Kingdom
- Member sinceSep 2014
Languages
English
I am a technologist with over 25 years of experience in cross-sector digital innovation and software development for startups and blue-chips. Currently, I lead GreenCode, focusing on developing AI solutions for climate and sustainability applications. I specialize in Generative AI, Local LLMs, and driving commercial impact through cutting-edge technology.
FAQ
What exactly is a "skill"?
A folder containing a SKILL.md file, in the open Agent Skills format. Your agent reads the short description at startup and loads the rest when a task matches, so it picks the skill up on its own rather than waiting to be pasted in.
How do I know it works?
You set the questions. We run each one with the skill loaded and without it, and hand you both answers side by side, so you can see what the skill adds before you deploy it.
Can you work from our own documents or recordings?
Yes, and the skill is better for it. Handbooks, process notes, past work, recorded talks and training sessions all make the skill more specific to how you work.
Which tools can load it?
Around forty, including Claude Code, Codex, GitHub Copilot, Cursor, Gemini CLI and VS Code. We give you the exact install path for each tool your team runs, so there is nothing to integrate.
Why not just ask an AI to write one?
For something small, do. A model writes a skill from what it already knows, which is the one thing your agent did not need telling. The value is in what it does not know: your conventions, how practitioners actually work, and proof on your own questions that the skill changed the answer.
Do you use our material to train anything?
No. It is used to build your skill and nothing else.
Do I own the result?
Yes. The skill is yours to edit, version and share. Third-party sources stay under their own licence: we build from them and cite them, and the skill is written in our own words.
