In the making

Your whole agent setup, in one place.

Skills, tools, plugins, automations and agent definitions.

We’ll only email you about early access.

How people build with AI

Community posts about the pieces behind better setups.

Illustrative layout – real posts to be added.

Clear instructions, the right tools and a little structure turn what an agent can do into work you can count on.

The five pieces

The parts most AI setups are made of, and what each one does.

Skills
Know-how your agent can reuse.
A skill is written guidance your agent follows for a certain kind of task, like reviewing code or writing release notes. Write it once and reuse it.
For example
A release-notes skill that uses your format and tone.
Tools
What your agent can actually use.
Tools let an agent run commands, call APIs or read from a service. MCP servers belong here too.
For example
A GitHub MCP server for opening and reviewing pull requests.
Plugins
Add-ons for the apps you already use.
Plugins extend the tools your setup can use and connect your agent to more systems.
For example
A plugin to search your docs from your agent.
Automations
Work that runs on its own.
Automations run steps on a schedule or when something happens: workflows, loops and graphs that chain tools and skills together.
For example
A nightly job that checks open pull requests and posts a summary.
Agent Definitions
Who the agent is, and where it stops.
An agent definition sets the role, system prompt, model and boundaries for one agent, so it behaves the same way every time.
For example
A reviewer agent that can read the repo but never push.