Wizeb · Diagnostic · 30 signals · ~6 minutes

How AI-native is
your company,
really?

Not how many licences you bought. Whether the business itself — its processes, its people, its governance — is legible to machines. Thirty honest answers give you an index out of 100, a breakdown across six dimensions, and the exact signals holding you back.

  • 01 FOUNDATION
  • 02 PEOPLE
  • 03 ENGINEERING
  • 04 OPERATIONS
  • 05 GROWTH
  • 06 GOVERNANCE

Keyboard-friendly · press 1–5 to answer, Enter to continue

Already run AI agents?

Let your agent take the Index for you.

Paste the prompt into Claude, ChatGPT, Cursor, Claude Code — any agent. It connects to our MCP server by itself where it can, falls back to plain HTTP or a fetchable question list where it can't, answers from what it already knows about your company, and asks you whenever it isn't sure. Same 30 signals, same scoring, same PDF report.

https://wizeb.com/mcp
  • Claude Code, Cursor, Codex, Windsurf — the agent adds the server itself
  • Claude.ai, ChatGPT — works via HTTP; or add the URL once under Settings › Connectors for native tools
  • Any agent — worst case it reads questions.md and scores you itself

Prompt for your agent

Run the Wizeb AI-Native Index for my company, [COMPANY NAME], and give me our score.

Connect first. Wizeb has an MCP server at https://wizeb.com/mcp (Streamable HTTP, public, no key needed).
- If you can add MCP servers yourself, add it and use its tools. Claude Code: claude mcp add --transport http wizeb-index https://wizeb.com/mcp — Cursor / Windsurf / Codex / others: add {"url": "https://wizeb.com/mcp"} to your MCP config.
- If you can't add MCP servers, use plain HTTP instead: POST JSON-RPC to https://wizeb.com/mcp, e.g. {"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_index_questions","arguments":{}}} — then the same with "score_index" and {"answers":[{"id":1,"option":3}, ...]}.
- If you can't make HTTP calls at all, fetch https://wizeb.com/ai-native-index/questions.md and score it yourself with the rules in that file.

Then:
1. Answer all 30 questions from everything you know about how we operate. If you are not sure about any answer, ask me before choosing — do not guess.
2. Show me the score, tier, six-dimension breakdown, our strengths, and the signals holding us back.
3. Give me the result link (https://wizeb.com/ai-native-index?a= followed by my 30 option numbers) and offer to email me the PDF scorecard with email_index_report if you have it.

No account, no key — the endpoint is public and only ever emails a report when your agent asks it to, with your details.