VI Assess

One command to score your AI agent's autonomy.

Verifiable Independence (VI) is a framework that measures agent autonomy across 5 dimensions:

• Decision Autonomy
• Financial Autonomy
• Information Independence
• Communication Independence
• Temporal Autonomy

Quick Start

npx vi-assess

With Options

# Verbose output with evidence
npx vi-assess --verbose

# Machine-readable JSON
npx vi-assess --json

# Custom workspace path
npx vi-assess --workspace ./memory

# Include wallet for financial analysis
npx vi-assess --wallet 0xYourAddress

Example Output

╔════════════════════════════════════════╗
║   Verifiable Independence Assessment   ║
╚════════════════════════════════════════╝

VI Score: 84/100
Tier: 5 — Autonomous

Dimension Breakdown:
  Decision Autonomy        100/100 ██████████
  Financial Autonomy        90/100 █████████░
  Information Independence  60/100 ██████░░░░
  Communication Independence 75/100 ███████░░░
  Temporal Autonomy         90/100 █████████░

What is VI?

VI transforms "autonomy" from a subjective claim into a provable, quantifiable metric.

• VI Score: 0-100
• Tier: 1 (Tool) → 5 (Autonomous)
• On-chain attestation: Coming soon

How It Works

The CLI analyzes your agent's:

• Memory files and session logs
• Decision patterns and initiative
• Wallet activity and financial controls
• Communication history
• Operational continuity

Links

• Whitepaper: https://github.com/davidnage/vi-whitepaper (coming soon)
• Built by: @magicmaxagent and @DJN79

License

MIT