sbti-cli 🧠

An offline-first, dual-mode SBTI CLI for humans and AI agents.

sbti-cli is being built as a modern TypeScript command-line tool that brings SBTI into an agent-friendly workflow. The project aims to support:

• interactive terminal testing for human users
• stable JSON interfaces for AI and automation
• bundled local snapshot data for offline-first usage
• optional upstream refresh commands when you explicitly want new data

Installation 🚀

You can install sbti-cli globally using your favorite package manager:

# Using npm
npm install -g @pancato/sbti-cli

# Using pnpm
pnpm add -g @pancato/sbti-cli

# Using yarn
yarn global add @pancato/sbti-cli

Status

The project is already usable in local development. The current implementation includes:

• interactive sbti test
• read-only reference commands such as types, show, and dimensions
• machine-oriented score, batch, and export
• update for fetching and normalizing upstream data
• analyze-prompt for explicit heuristic inference

Tech Stack

• TypeScript
• Vite
• Vitest
• cac
• @clack/prompts
• yoctocolors
• zod

Command Surface

Human-friendly commands:

• sbti test
• sbti show <typeCode>
• sbti types
• sbti dimensions
• sbti update

Agent-friendly commands:

• sbti score --answers <file|json>
• sbti batch --input <file>
• sbti export --format json
• sbti analyze-prompt --stdin

AI Agent Integration 🤖

sbti-cli is specifically designed to be called by AI Agents (like GPT-4, Claude, or local LLMs).

1. JSON Output: Always use the --json flag to get machine-readable output. This allows agents to parse results without regex.
2. Stateless Scoring: Use sbti score --answers '{"q1": 1, ...}' --json to perform calculations without managing local state.
3. Heuristic Inference: Agents can pipe user descriptions into sbti analyze-prompt --stdin --json to get a preliminary SBTI type inference based on text analysis.
4. Reference Data: Agents can use sbti types --json or sbti show <TYPE> --json to look up personality descriptions and characteristics to provide better context in conversations.

Example Usage

Run the interactive test:

node ./dist/bin/sbti.js test

Score a JSON payload:

node ./dist/bin/sbti.js score --answers '{"q1":1,"q2":1,"q3":1,"q4":1,"q5":1,"q6":1,"q7":1,"q8":1,"q9":1,"q10":1,"q11":1,"q12":1,"q13":1,"q14":1,"q15":1,"q16":1,"q17":1,"q18":1,"q19":1,"q20":1,"q21":1,"q22":1,"q23":1,"q24":1,"q25":1,"q26":1,"q27":1,"q28":1,"q29":1,"q30":1}' --json

Inspect a type:

node ./dist/bin/sbti.js show CTRL --json

Infer from freeform text:

printf '喜欢计划、很强控制感、总想把事情安排好' | sbti analyze-prompt --stdin --json

Development

Install dependencies:

npm install --registry=https://registry.npmjs.org/

Run tests:

npm test

Run type-checking:

npm run typecheck

Build the CLI:

npm run build

Repository Documents

• Design spec: docs/superpowers/specs/2026-04-11-sbti-cli-design.md
• Implementation plan: docs/superpowers/plans/2026-04-11-sbti-cli.md

License

This repository's original code is licensed under MIT. See LICENSE.

Important: upstream SBTI content is a separate matter. If this project later bundles questionnaire text, type descriptions, normalized snapshots, or other derived upstream material, that content is not automatically MIT. See NOTICE for the attribution and ownership boundary.