TIBET Core

The Linux of AI Provenance

A minimal, embeddable provenance engine for any device. From microcontrollers to cloud servers.

What is TIBET?

Transaction/Interaction-Based Evidence Trail

TIBET captures the four dimensions of every AI action:

Installation

Python

pip install tibet-core

JavaScript/Node.js

npm install tibet-core

Rust

[dependencies]
tibet-core = "0.1"

Quick Start

Python

from tibet_core import TibetEngine

engine = TibetEngine()

# Create a provenance token
token = engine.create_token(
    token_type="action",
    erin="User requested translation",
    eraan=["model_v1", "tokenizer_v2"],
    eromheen='{"env": "production"}',
    erachter="Fulfilling user request",
    actor="agent_001"
)

print(f"Token ID: {token.id}")
print(f"Valid: {token.verify()}")
print(f"JSON: {token.to_json()}")

JavaScript

import { TibetEngine } from 'tibet-core';

const engine = new TibetEngine();

const tokenJson = engine.create_token(
    "action",
    "User requested translation",
    JSON.stringify(["model_v1", "tokenizer_v2"]),
    '{"env": "production"}',
    "Fulfilling user request",
    "agent_001",
    null  // no parent
);

const token = JSON.parse(tokenJson);
console.log(`Token ID: ${token.id}`);
console.log(`Valid: ${engine.verify(tokenJson)}`);

Rust

use tibet_core::TibetEngine;

let engine = TibetEngine::new();

let token = engine.create_token(
    "action",
    "User requested translation",
    &["model_v1", "tokenizer_v2"],
    r#"{"env": "production"}"#,
    "Fulfilling user request",
    "agent_001",
    None,
);

assert!(token.verify());

Chaining Tokens

Create audit trails by linking tokens:

# Parent action
request = engine.create_token(
    token_type="request",
    erin="Translate 'hello' to Dutch",
    eraan=[],
    eromheen='{"user": "alice"}',
    erachter="User wants translation",
    actor="user_001"
)

# Child response (linked to parent)
response = engine.create_token(
    token_type="response",
    erin="Hallo",
    eraan=["gpt-4"],
    eromheen='{"latency_ms": 150}',
    erachter="Translation completed",
    actor="ai_agent",
    parent_id=request.id  # Chain link!
)

print(f"Parent: {request.id}")
print(f"Child parent_id: {response.parent_id}")  # Same as request.id

Why TIBET?

Regulatory Compliance

• GDPR Article 22 (automated decision-making)
• EU AI Act (transparency requirements)
• SOC 2 / ISO 27001 (audit trails)

6G Ready

• Designed for AI-native networks (ITU IMT-2030)
• Referenced in IETF 6G AI agent drafts
• Minimal footprint for edge devices

Cryptographically Secure

• Ed25519 signatures
• Tamper-evident chains
• Verifiable without central authority

Size Comparison

IETF Standardization

TIBET is being standardized at the IETF:

• draft-vandemeent-tibet-provenance - Evidence Trail Protocol
• draft-vandemeent-jis-identity - JTel Identity Standard

W3C Alignment

TIBET aligns with W3C standards:

• Verifiable Credentials 2.0 - Token structure compatible
• Decentralized Identifiers (DIDs) - Actor identification
• JSON-LD - Semantic context in EROMHEEN

Architecture

┌─────────────────────────────────────────────────────────────┐
│                    TIBET ECOSYSTEM                          │
├─────────────────────────────────────────────────────────────┤
│   ┌─────────────────┐                                       │
│   │  tibet-core     │  ← 50KB Rust, minimal deps           │
│   │  (the kernel)   │    create_token, verify, chain       │
│   └────────┬────────┘                                       │
│            │                                                │
│   ┌────────┴────────┬──────────────┬──────────────┐        │
│   ▼                 ▼              ▼              ▼        │
│ tibet-c        tibet-wasm    tibet-python   tibet-js       │
│ (embedded)     (browser)     (PyPI)         (npm)          │
└─────────────────────────────────────────────────────────────┘

Credits

• Specification: Jasper van de Meent (Humotica)
• Implementation: Root AI (Claude) & Jasper van de Meent
• License: MIT OR Apache-2.0

Links

• Humotica
• IETF TIBET Draft
• IETF JIS Draft
• PyPI
• npm

"The Linux of AI Provenance" - Making audit trails as universal as the kernel.