We build scoped custom agents. This page is not a shop. It exists to show one skill in full detail, because a skill is the unit an agent is actually assembled from, and seeing a real one makes the difference between an agent and a chat window concrete.
What follows is a skill definition for Claude Code, not an app and not a hosted service. A skill like this runs on your own machine and your own Claude subscription. We are showing it because it is a real, finished example of the smallest unit we work in. If you want a capability like this built around your own operation, that is a scoped custom build, and the way to start is a conversation rather than a checkout.
Why we show the whole thing. Most capability claims are a sentence in a brochure. A skill is a written procedure with stated triggers, stated checks and a stated refusal, which means it can be read and judged before anyone trusts it. That is the standard we hold our own work to, so it is the standard we are willing to show.
What a skill is, in one line. A written, scoped capability an agent loads when the work calls for it: when to use it, what it must check first, and what it refuses. Closer to a checked procedure than to a personality trait, which is why an agent's capability list is an inventory rather than a claim.
What this page is not. Not a shop and not a catalogue. We sell one thing, a scoped custom agent build, and skills are shown here only as reference so the work is legible before you talk to us. Nothing on this page is purchasable.
Why only one is shown. We would rather show one skill we have actually packaged and run than a page of examples that do not exist yet. More will appear here as they are genuinely finished.
A ready-made skill covers a job somebody has already defined. When the operation is yours and nobody has built it yet, that is a scoped custom build: we agree the deliverable, the date and the revisions in writing, build it, and hand it over for you to run.