Installation

Get started with CCF

CCF is available as a Rust crate on crates.io. It compiles to no_std targets including embedded ARM and WASM appliance tiles.

Cargo install

Cargo.toml

[dependencies]
ccf-core = "0.1"

Or via the CLI: cargo add ccf-core

Feature flags

serdeEnable Serialize/Deserialize on all public types for session persistence (CcfSegSnapshot).

stdOpt into std for environments that have it. Default is no_std + alloc.

python-ffiExpose a PyO3 FFI layer. Requires maturin build.

Cargo.toml — with serde

[dependencies]
ccf-core = { version = "0.1", features = ["serde"] }

Quick start

src/main.rs

use ccf_core::{
    accumulator::CoherenceField,
    phase::Personality,
    vocabulary::{ContextKey, SensorVocabulary},
};

// 1. Define your sensor vocabulary
#[derive(Clone, PartialEq, Eq, Hash, Debug)]
struct MySensors { proximity: u8, light: u8 }

impl SensorVocabulary<2> for MySensors {
    fn to_feature_vec(&self) -> [f32; 2] {
        [self.proximity as f32 / 3.0, self.light as f32 / 2.0]
    }
}

fn main() {
    let mut field = CoherenceField::new();
    let personality = Personality::default();

    // 2. Observe an interaction
    let sensors = MySensors { proximity: 2, light: 1 };
    let key = ContextKey::<MySensors, 2>::new(sensors);

    for tick in 0..20u64 {
        field.positive_interaction(&key, &personality, tick, false);
    }

    // 3. Read the social phase
    let coherence = field.context_coherence(&key);
    println!("Coherence: {:.3}", coherence); // e.g. 0.847
}

API docs on docs.rs →Examples on GitHub →

Reference implementation

mBot2 — the $50 robot

The canonical CCF demo runs on an mBot2 connected via Bluetooth LE. Eight sensor dimensions drive a CCF brain — the robot develops measurably different social phases per context with no training data.

Cargo.toml — mBot2 target

[dependencies]
ccf-core = { version = "0.1", features = ["serde"] }

# Optional: BLE transport
btleplug = "0.11"

Source: github.com/Hulupeep/CCF — see crates/mbot-companion

Python FFI

Use CCF from Python

The python-ffi feature exposes CCF via PyO3. Build with maturin, install as a local wheel, and drive the coherence field from Python.

Build & install

pip install maturin
maturin develop --features python-ffi

Python usage

import ccf_core

field = ccf_core.CoherenceField()
personality = ccf_core.Personality()

# Drive positive interactions
for tick in range(20):
    field.positive_interaction([0.67, 0.5], personality, tick, False)

coherence = field.context_coherence([0.67, 0.5])
print(f"Coherence: {coherence:.3f}")

crates.io/crates/ccf-core docs.rs/ccf-core github.com/Hulupeep/ccf-core