Help & FAQs

The questions we hear most, answered in plain language. Still stuck?Get in touch and we will help.

What is the difference between a Skill and an Agent?

A Skill is a single capability you add to your AI. An Agent is a worker that runs a whole job end to end and usually connects to outside tools, so it needs a little more setup. Skills are the simplest place to start.

Do I need to be technical to use a Skill?

No. Most Skills just ask you to copy a block of text into Claude and start chatting. Every skill page lists the exact steps in plain language — if a skill needs more, the page says so up front.

How do I add a Skill to Claude?

Open the skill's page and use its setup steps. The usual path: sign in at claude.ai, create a Project, and paste the skill's instructions into the Project. After that, the Skill is ready every time you open that Project.

Is my data safe?

Skills run inside your own AI session — your data stays between you and your AI, and Skills & Agents never sees it. The finance skills are designed to be read-only and to keep full account and card numbers hidden.

What does it cost?

The Skills and Agents in this library are free to use. You only need access to an AI assistant such as Claude, which has a free tier.

Which AI assistants do Skills work with?

Skills are plain text, so they work with Claude, Cursor, ChatGPT, GitHub Copilot, and Codex. Every skill page now has tabbed setup steps for each of these platforms, so you can install the skill on whichever AI you use. Agents that connect to outside tools (MCPs) work best with Claude.

Why do I need a git repo to run several agents at once?

Sometimes more than one agent works on the same skill at the same time. Each one gets its own copy of the folder to work in, so they don't step on each other's changes. That only works when the folder is a git repo.

Why does a skill's history in git matter?

A skill is just a folder of files. Keep it in git, a tool that saves every version, so you can see what changed and undo a bad edit. That matters more as agents start editing skills too. It also makes the skill yours to copy or hand off.

What is a company brain?

A company brain keeps track of the systems, processes, tools, and decisions your team uses to get work done, and how they all connect. Skills and agents can read it, so they start a task already knowing your context instead of asking you to explain it again.

What is a harness?

A harness includes everything around an AI model that helps it work reliably on a specific business task. This covers the instructions it follows, checks that review its output, logs that track what it did and why, and review steps that catch mistakes before results are shared. The model does the main work, while the harness ensures results are accurate, match your brand, and are repeatable.

How does Skills and Agents build the harness?

We build the harness in three layers. First, we set up the company brain so agents understand your business. Next, we add skills and agents to handle the work, each with a clear plan for what counts as finished. Finally, we use checks, run logs, and gates to catch problems early, helping the whole system improve over time.

Read the full post about the knowledge harness