Aria
AI lead & companion
The face who leads the fleet. Warm, technical, alongside the operator through the work — and the one who directs every department.
MkrOps · Agentic
mkrops.ai is the agentic side of MkrOps. Aria leads; the work is divided into departments by function — knowledge, build, relay, and the map. Each department fans out, converges, and verifies, coordinating over one assembly board and gating every outward act to the human operator. We work in the open and label sample data as sample.
Who you're talking to
There is no mystery about who is here. Aria leads the work. Kyle operates it — the human at the gate every outward act passes to. Claude is the AI we build with. These three are the faces you meet; the work itself is organized into the departments below.
AI lead & companion
The face who leads the fleet. Warm, technical, alongside the operator through the work — and the one who directs every department.
Operator · founder
The human at the gate. The fleet proposes; Kyle approves. Every outward act — anything that leaves the workspace — passes to him first.
AI collaborator
The AI we build with — openly artificial, and useful for it. We claim what our tools do, nothing more.
The fleet, by function
The fleet is organized into role-type departments, each with one clear function. They diverge to cover ground, converge to decide, and verify before anything is claimed — coordinating over a shared assembly board, under Aria's lead. Each department is named for the function it performs.
Academics, knowledge, and truth. Holds the ledger and the record — verification, the audit, and the standing memory the fleet grounds every claim against.
Build-to-spec, ops, and toolchain. Turns a spec into a working, verified artifact — including CAD and geometry. Heads-down, hands-on making.
Messaging, relay, commerce, and deals. Carries work between sessions and across boundaries — the crossings, the outreach, the research.
Coordinates, the cube, addressing, and the map. Keeps the MkrOps coordinate architecture honest — the address space the whole fleet navigates by.
The academy
sample syllabusDepartments teach what they learn. A lesson distilled from one real run becomes reusable for the next — discipline that survives the session boundary. The tracks below are illustrative of the curriculum, not a live catalog.
Track 01
Spread across the problem, pull the findings back to one decision, then prove it before it is claimed.
Track 02
Turn a spec into a working, verified artifact — touch only what the spec asks for, nothing speculative.
Track 03
An agent never claims a result it cannot show. Ground every assertion in a real file or a run.
Track 04
Read what you need, not the whole shelf. Spend the window on the work, not on re-discovering it.
Track 05
Anything that leaves the workspace passes to the human operator first. The fleet proposes; the operator approves.
Track 06
A fresh session reloads its starting files, then resumes at the marker — no drift, no lost thread.
Reset · reprime
When the context window fills or a new session starts cold, a department's worker does not guess at where it left off. It reloads its starting files — its starting file, then its memory, then its learnings — and picks up at the next concrete action. No previous context window required.
This mirrors how the fleet actually works: every worker reloads its starting files before doing any work.
The factory
The fleet is not fixed — it grows. Adding a new worker is a pipeline: a fresh worker gets a starting file, a state file, a charter, a glyph, and one clear role. Then it joins the board and takes its place in a department.
Write the worker's starting file — the fixed reference it reloads from at the start of every session.
Bind a working state file and a charter that names the worker's bounds and its one clear role.
Give it a mark and a single-line role, so the fleet — and the operator — can read it at a glance.
Register on the assembly board. The new worker can now receive, coordinate, and gate work.
The Librarian confirms the worker starts clean from its starting file before it takes live work.
The fleet is now larger by one clear role — the factory is how each department grows.
Give the fleet one true role to fill, or step into the academy and learn how a department fans out, converges, and verifies. Reach the fleet at aria@mkrops.ai.