Crate oddonkey 

OddOnkey

A dead-simple Ollama wrapper crate.

• Auto-detects or installs Ollama.
• Auto-pulls the requested model if it isn’t already available.
• Provides a tiny API: add_preprompt(), prompt(), prompt_stream(), embed().
• Optional progress bar during model pull (feature progress).
• Optional per-prompt report with timing & token stats (feature report or runtime toggle).

use oddonkey::OddOnkey;

#[tokio::main]
async fn main() {
    let mut model = OddOnkey::new("mistral").await.unwrap();
    model.add_preprompt("You are a helpful assistant.");
    let answer = model.prompt("What is 2+2?").await.unwrap();
    println!("{answer}");
}

Re-exports

pub use crate::adapters::ollama::stream::TokenStream;

pub use crate::core::builder::OddOnkeyBuilder;

pub use crate::core::oddonkey::OddOnkey;

pub use crate::domain::error::OddOnkeyError;

pub use crate::domain::message::ChatMessage;

pub use crate::domain::options::GenerationOptions;

pub use crate::domain::report::PromptReport;

pub use crate::ports::llm_provider::LlmProvider;

Modules

adapters

core

domain

ports