attuned-core

Core types and traits for Attuned - a Rust framework for representing human state as interpretable vectors and translating them into interaction constraints for LLM systems.

Overview

Attuned produces context, not actions. It represents user state across interpretable dimensions (axes) and translates that state into guidelines for LLM interactions.

Core Types

• [StateSnapshot] - A point-in-time capture of user state
• [PromptContext] - Translated guidelines for LLM conditioning
• [Translator] - Trait for converting state to context

Example

use attuned_core::{StateSnapshot, Source, RuleTranslator, Translator};

// Create a state snapshot
let snapshot = StateSnapshot::builder()
    .user_id("user_123")
    .source(Source::SelfReport)
    .axis("warmth", 0.7)
    .axis("cognitive_load", 0.9)
    .build()
    .unwrap();

// Translate to prompt context
let translator = RuleTranslator::default();
let context = translator.to_prompt_context(&snapshot);

// Use guidelines in your LLM system prompt
for guideline in &context.guidelines {
    println!("{}", guideline);
}