Work with me · Consulting
Three ways to work together, each with a published price and a fixed scope — because “contact us for pricing” wastes a week of everyone's time. Every engagement runs on the same method as the Agent-Ready Index, applied to your system instead of the public 37.
Write once, hear back within 2 business days.
Straight answer on availability: consulting is limited and by enquiry right now — there is no booking calendar, and the work below is taken on case by case. The current focus is the product line, so if you would rather not wait on my schedule, start with the products, which package the same method for you to run yourself.
01 — The offers
Start with the audit if you want the diagnosis before committing to treatment. The 90-day sprint implements the roadmap; the fractional arrangement is for teams that need the leadership, not just the deliverables.
01 · 2 weeks, fixed
$3,500
Your design system scored against the five signals of the Agent-Ready Index — MCP, llms.txt, W3C/DTCG tokens, component registry, Code Connect — plus token architecture and docs machine-readability. You get the scored evaluation, first-party evidence for every finding, and a sequenced remediation roadmap your team can execute without me.
02 · fixed scope, 6–12 weeks
from $15,000
The audit roadmap, implemented: llms.txt, DTCG token migration, a shadcn-style component registry, an MCP server over your system, and Code Connect for your most-used components. Scoped and priced before we start.
03 · 1–2 days a week · 2 slots
$4,000–8,000/mo
Senior design-systems leadership without the headcount: governance, contribution model, adoption measurement, and the agent-readiness work, embedded in your team. Capped at two concurrent engagements.
There is no booking calendar on purpose — with consulting deprioritised, a self-serve slot picker would promise availability I can't guarantee. The button opens your email client with the subject pre-filled. Same inbox, same 2-business-day reply.
02 — How the audit works
The audit scores your system against the same five signals the Agent-Ready Index tracks across 37 public systems — plus token architecture and docs machine-readability. Every finding cites evidence; every recommendation is sequenced so your team can execute it in order.
Signal 01
Can an agent query your system directly — components, tokens, usage — or does it have to scrape docs written for humans?
Signal 02
Is there a machine-readable map of your documentation, or do agents burn context guessing which page matters?
Signal 03
Are your tokens in the W3C community format that tools and agents can consume, or locked in a proprietary export?
Signal 04
Can components be pulled into a codebase by name — shadcn-style — instead of copy-pasted from a docs site?
Signal 05
Does design-tool output map to your real components, or does every handoff regenerate divergent code?
The receipt
Same rubric, public data, quarterly re-audits. Judge the methodology before you buy it.
03 — Why me
Judgment is only worth what it produces in public. Everything I'd bring to your engagement already exists on this site, dated and sourced.
04 — Intake
Five fields, two minutes. I read every message myself and reply within 2 business days — usually with two or three questions, sometimes with “you don't need me, here's the chapter to read.”
Prefer plain email? Write to kir.zhukovsky@gmail.com directly.
