Connect DimensionKit to Claude, ChatGPT or Cursor in a minute. Your assistant already knows your framework: ask, and it writes the scored questions, sets the tiers and shows you a preview before anything goes live. Your visitors can even answer with their own AI.
Every visitor gets a personalised result. You get their email, their answers, and a reason to follow up.
Works with Claude, ChatGPT, Cursor and any assistant that supports remote MCP.
Open beta · free · no credit card. See pricing.
AI on both sides
DimensionKit plugs into the assistants you and your audience already use. The assistant brings the context; DimensionKit brings the scoring, the results page and the leads.
Turn the RevOps framework we worked on into a scorecard.
1You build
It has your notes, your documents and your past chats. Ask for a scorecard and it creates the dimensions, writes the scored questions, sets the tiers and shows you a preview before anything goes live.
Answer this scorecard from our CRM and docs: …/revops-maturity.md
2They answer
Every published scorecard has a plain-text version an assistant can read. It answers from real evidence — CRM, docs, dashboards — asks when it isn't sure, and hands back a link. Visitors review every answer, then submit.
How visitors answer with an assistantWho scored lowest on Pipeline this month?
3You follow up
Your assistant reads the submissions and their per-dimension scores. Which leads are weak on exactly what you sell? Who should hear from you this week? Just ask.
Set up in a minute
Paste the server URL into your assistant's connector settings. It sends you to sign in and asks what it may do — read only, or read and write. That is the whole setup.
MCP server URL
https://app.dimensionkit.com/mcpIn Claude, ChatGPT, Cursor or any client that supports remote MCP servers, add it as a custom connector.
No account yet? Start anyway — you can sign up on the way. Choose read only, or read and write.
“Build me a scorecard that assesses technical leadership.” Your assistant interviews you, builds it and sends you a preview link.
Connecting brings up a consent screen with two choices: read, or read and write. A read-only connection cannot change anything. Every connection acts as you and reaches only your own scorecards.
Change a published scorecard whenever you want. Submissions already collected keep the scores and the wording they were given — your history doesn't shift under you.
A Connected apps page in the admin shows everything currently connected to your account, and disconnecting ends its access immediately.
Scoring stays exactly as you defined it — the same answers always produce the same result. An assistant is a way to build your scorecards and read the results, not a way to score the people taking them. And none of it is required: everything here can still be built by hand in the admin.
What you get
Generic ebooks get downloaded and forgotten. An assessment gives every visitor a personal result — and gives you their email, their scores, and a clear reason to follow up.
You get the email, the scores, and the context. Every submission is tied to real answers — so you know who's ready to talk and who needs nurturing.
A radar chart, per-dimension tiers, and tailored insights — generated from each visitor's answers. Same assessment, different report for everyone.
One script tag drops your assessment into any approved site. Locked to your domains only — every other origin is blocked.
What your visitor sees
Whether your assessment lives on a landing page or embeds inside your marketing site, this is the experience every visitor walks through.
Multiple choice, one at a time, with progress. Mobile-friendly out of the box.
With whatever fields you choose — name, company, team size, custom. Decide whether email comes before or after questions.
Radar chart, per-dimension tiers, insight copy. Shown on screen and emailed within minutes.
What you get back
Every submission lands in one place — with scores, tiers, source domain, and any custom fields. Build new assessments without writing code.
Email, company, scores, status, source domain — searchable, filterable, exportable. Nothing leaves your admin.
Dimensions, questions, tiers, branding, email capture — edit in the admin, save, done. No deploys. Or connect an AI assistant and talk it through instead.
Use cases
If you can express your method as dimensions with questions and tier-based outcomes, you can ship it as an assessment. We call each one a scorecard.
Show prospects where they stand on the framework you teach. A clean opener for a sales conversation.
Multi-dimensional assessments with tier-based insights — for org maturity, team practices, technical readiness.
Replace the generic PDF with a scored, interactive experience. Email capture and delivery built in.
Let clients check their own progress between engagements. Embed on a client portal with its own allowed origin.
Use it as pre-work — qualify applicants and arrive at the first session knowing where each person is.
Bundle your framework as a standalone product. White-label colors, your CTA, your follow-up copy.
Features
Every published scorecard also has a plain-text version an assistant can read. Visitors' assistants answer from evidence, flag what they can't tell, and hand back a prefilled link to review and submit.
Unlimited dimensions, any number of questions per dimension, custom score values, and configurable tier thresholds with per-tier insight copy.
Paste one URL into any client that supports remote MCP servers — no API keys to create. Your assistant can build and edit scorecards from your description, and read submissions with their scores. You approve read-only or read-and-write, and can disconnect it from the admin at any time.
Per-dimension scores rendered as a radar chart with tier badges and insight copy — readable in light and dark mode, accessible to screen readers.
Manage assessments, view submissions, and configure embed allowed-origins through a built-in admin UI. Multi-user with roles.
Submissions are queued and emailed via SMTP with configurable from and reply-to addresses. Visitors can update their email with a tokenised link.
Brand colors applied at runtime via CSS custom properties. Multiple themes available: default, obsidian, playground, brutalist. Light and dark with system detection.
Each assessment has its own list of allowed embed origins. The embed route sends a route-specific frame-ancestors CSP header — every other origin gets 403.
Also included
Yours to embed
Share the standalone link, or drop your assessment straight into your marketing site with a single script tag — locked to the origins you approve, so nobody else can host it.
The loader creates a responsive, sandboxed iframe. The embed route sends a per-assessment frame-ancestors CSP header — every other origin gets 403.
<div id="scorecard"></div>
<script
src="https://app.dimensionkit.com/embed.js"
data-dk-scorecard="your-scorecard-slug"
data-dk-target="#scorecard"
></script>Open beta — free
DimensionKit is free during open beta — two scorecards, 100 submissions a month, and no credit card.