eval-magic 0.4.0

One-stop CLI for running skill evals — measure whether an agent skill actually shifts behavior.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

//! Integration tests for the CLI surface, driving the built `eval-magic`
//! binary. These pin the command tree and dispatch behavior; per-command
//! behavior lives with each subcommand's submodule.
//!
//! Split by subcommand into submodules (file-length guideline); shared helpers
//! ([`skill_eval`](helpers::skill_eval), [`canonical_root`](helpers::canonical_root))
//! live in [`helpers`], single-use helpers stay with their group.

mod helpers;

mod aggregate;
mod basics;
mod grade;
mod grade_models;
mod guard;
mod init;
mod package;
mod stray_writes;
mod workspace;