Tenrung in Pi

Questions it answers

Tenrung scores how far a team has come from paying for AI seats to working with AI every day. The user scores 10 lines from 0 to 3, so the total runs 0 to 30.

The lines cover ownership (a named AI owner with hours for it, department reps meeting every 2 weeks, a 1-page policy), learning (monthly live training on the team's own work, 3 training tracks, a recorded library checked every month) and the loop (a use-case list ranked by hours saved, wins posted company-wide, weekly blocked AI time, hours saved shown to leadership).

The answer gives the band, the 3 first moves from the weakest lines, and a 90-day plan: month 1 sets up the champion, the policy, the library and office hours; month 2 ships 2 workflows, times them, then ships 2 more; month 3 runs a 2-week hackathon with a demo day judged 50% on hours saved, 30% on how many would use it and 20% on reuse. Each step is marked start, tighten or keep from the scores.

It also returns a role card for the AI champion: who fits, the 5 recurring jobs and about 5 hours a week. The score is a self-assessment, not an audit. Nothing is stored; the share link holds only the 10 numbers and an optional team name.

What it can do

• Scores 10 lines from 0 to 3 and returns a total out of 30 with one of 4 bands
• Picks the 3 first moves from the weakest lines
• Builds a 90-day plan with every step marked start, tighten or keep
• Returns the AI champion role card with 5 jobs and the weekly hours
• Gives a shareable result link that holds only the 10 numbers

Tools

• score_ai_native_team, Score how AI-native a team is. Score a team or company on the 10-line AI-native scorecard and get a 90-day plan. Each line is scored 0 to 3 by the user: a named AI owner with hours for it, department reps meeting every 2 weeks, a pinned 1-page AI policy, monthly live training on their own workflows, training in 3 tracks, a recorded session library checked monthly, a use-case list ranked by hours saved, wins posted company-wide, weekly blocked AI time, and hours saved measured and shown to leadership. Ask the user for each of the 10 scores first (the input schema says what 0, 1, 2 and 3 mean on each line); never guess one. Returns the total out of 30, the band (You bought seats, A start no engine, Almost there, AI native), the 3 first moves, a 90-day plan in 3 months with each step marked start, tighten or keep, the AI champion role card, and a share_url to a public result page. Nothing is stored; the link holds only the 10 numbers and the optional team name.

Add Tenrung in Pi

Pi does not speak MCP on its own. The pi-mcp-adapter extension (these steps match version 2.34.0) loads remote servers and keeps their schemas out of the prompt until you ask. Install it once, then restart Pi:

pi install npm:pi-mcp-adapter

Add the server to a shared MCP file. Use .mcp.json in the project if only that project needs Tenrung, or ~/.config/mcp/mcp.json if you want it in every project. Pi also reads ~/.agents/mcp.json. A Pi-only override lives in ~/.pi/agent/mcp.json or .pi/mcp.json and wins over the shared file.

{
  "mcpServers": {
    "tenrung": {
      "url": "https://tenrung.openkrill.app/mcp",
      "lifecycle": "eager"
    }
  }
}

The url is enough. Do not set command: that field is for a local process. No API key header. lifecycle is eager so Pi connects at startup and can search tool names before the first call. A config that omits it still works: the first call connects. After restart, run /mcp and confirm the server is listed. Ask: Score how AI-native my team is and give me a 90-day plan. The adapter should offer score_ai_native_team.

Leave directTools off unless you want those schemas in every turn. With a handful of tools the cost is small. The default proxy is the better fit: search for the tool, then call it. /mcp disable tenrung writes a project override in .pi/mcp.json without editing the shared file. Run /reload after that change.

Pi will not find this server by scanning Cursor or Claude config unless you run /mcp setup and import it. Putting the url in .mcp.json is the direct path and the one this page tests against. The server is the same public endpoint the other clients use, with the same limits and the same privacy rules.

If it does not work

A call checked on 2026-10-11

A made-up team with an owner and a policy but no measured hours. The tool reads only the 10 numbers and keeps nothing. The request below was posted to https://tenrung.openkrill.app/mcp as tools/call. HTTP 200. Source: the Tenrung server, read 2026-10-11.

{
  "method": "tools/call",
  "params": {
    "name": "score_ai_native_team",
    "arguments": {
      "scores": {
        "owner": 3,
        "reps": 2,
        "policy": 3,
        "training": 2,
        "tracks": 1,
        "library": 2,
        "backlog": 1,
        "wins": 2,
        "focus_time": 1,
        "hours_saved": 0
      },
      "team_name": "Northwind"
    }
  }
}

• lines. 10 returned
• lines title. One named person owns AI adoption, with hours for it in their week
• first moves. 3 returned
• first moves title. Hours saved are measured and shown to leadership
• plan. 3 returned
• plan name. Set up

Repeat the call yourself if you need a newer reading. Cached answers expire. A rate limit is not a result: wait and try again. Nothing in the call is a ranking, a filing, or advice.

Using it in production

Nothing to install and nothing to run: no package, no Python or Node.js version and no environment variable. Tenrung runs on our servers, and your client only needs the URL https://tenrung.openkrill.app/mcp.

Use it in other apps

• Tenrung in ChatGPT
• Tenrung in Claude
• Tenrung in Cursor
• Tenrung in VS Code

Frequently asked questions

Words on this page

• MCP (Model Context Protocol) definition
• Streamable HTTP transport definition
• Tool primitive definition
• Rate limit policy definition

Source: plugin listing for tenrung in this repository, and a live tools/call on 2026-10-11. As of 2026-10-11.