agentcarousel 0.5.7

Unit tests for AI agents. Run behavioral tests in CI, score with an LLM judge, and export signed evidence your auditors accept.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

//! Human-readable and machine-readable **output**: terminal tables, JSON, JUnit, persisted
//! history (SQLite), and [`diff_runs`] / [`print_diff`] for comparing two runs.

mod diff;
mod history;
mod junit;
mod terminal;

pub use diff::{diff_runs, print_diff};
pub use history::{fetch_run, list_full_runs, list_runs, persist_run, HistoryError, RunListing};
pub use junit::print_junit;
pub use terminal::{print_terminal, print_terminal_summary};

pub fn print_json(run: &agentcarousel_core::Run) {
    let payload = serde_json::to_string_pretty(run).unwrap_or_else(|_| "{}".to_string());
    println!("{payload}");
}