chaos_theory

chaos_theory is a modern Rust property-based testing and structure-aware fuzzing library.

You drive tests using Source to get structured pseudo-random values and control flow, chaos_theory ensures that this exploration is efficient and any failures found are automatically minimized.

Quickstart

use chaos_theory::{check, make::string_matching};

#[test]
fn slug_and_id_roundtrip() {
    check(|src| {
        let slug = src.any_of("slug", string_matching("[a-z0-9]+(-[a-z0-9]+)*", true));
        let id: u32 = src.any("id");

        let s = format!("{slug}-{id}");
        let (slug_parsed, id_parsed) = s
            .rsplit_once('-')
            .map(|(s, i)| (s, i.parse::<u32>().unwrap()))
            .unwrap();

        assert_eq!(slug_parsed, slug);
        assert_eq!(id_parsed, id);
    });
}

When a failure happens, chaos_theory prints a CHAOS_THEORY_REPLAY=... string you can use to reproduce the case.

Highlights

• Property testing and structure-aware fuzzing in one library
• Efficient state space exploration:
  • bias towards small values and edge cases
  • structural mutations and crossover
  • example-guided generation
  • built-in swarm testing
• Macro-free, imperative API (with optional derive macro for Arbitrary)
• Zero unsafe code and zero required dependencies

Documentation

https://docs.rs/chaos_theory

Status

Beta – chaos_theory works well and is useful, but does not guarantee API stability.

Notable gaps:

• Proper recursive generators

Contributing

• Feedback – bug reports, issues, proposals – is welcome
• Pull requests require prior discussion before being open
• All AI usage must be disclosed

License

MPL-2.0, see LICENSE.