review AGENTS.md manually

Author: basvanwesting

AGENTS.md

@@ -176,12 +176,12 @@ with `call_repeatedly` for genomes >20 genes.
 | Genotype | Compatible Crossovers | Recommended |
 |---|---|---|
 | `BinaryGenotype` | All | `CrossoverUniform` or `CrossoverSinglePoint` |
-| `ListGenotype<T>` | All | `CrossoverUniform` |
-| `MultiListGenotype<T>` | All | `CrossoverUniform` |
+| `ListGenotype<T>` | All | Any (e.g. `CrossoverUniform`, `CrossoverSinglePoint`) |
+| `MultiListGenotype<T>` | All | Any (e.g. `CrossoverUniform`, `CrossoverSinglePoint`) |
 | `UniqueGenotype<T>` | `CrossoverClone`, `CrossoverRejuvenate` ONLY (others are compile errors) | `CrossoverClone` |
 | `MultiUniqueGenotype<T>` | Point-based + `CrossoverClone`, `CrossoverRejuvenate` (gene-based are compile errors) | `CrossoverSinglePoint` |
-| `RangeGenotype<T>` | All | `CrossoverMultiPoint` |
-| `MultiRangeGenotype<T>` | All | `CrossoverSingleGene` |
+| `RangeGenotype<T>` | All | Any (e.g. `CrossoverUniform`, `CrossoverSinglePoint`) |
+| `MultiRangeGenotype<T>` | All | Any (e.g. `CrossoverUniform`, `CrossoverSingleGene`) |
 
 **Compile-time safety**: `UniqueGenotype` does not implement `SupportsGeneCrossover`
 or `SupportsPointCrossover`, so incompatible crossovers are **compile errors**.
@@ -249,7 +249,7 @@ For range/float genotypes (>50 genes, see Troubleshooting for tuning):
 ```rust
 // also requires: genotype, fitness, target_population_size, ending condition
 .with_select(SelectTournament::new(0.5, 0.02, 4))
-.with_crossover(CrossoverMultiPoint::new(0.7, 0.8, 3, false))
+.with_crossover(CrossoverUniform::new(0.7, 0.8))
 .with_mutate(MutateMultiGene::new(10, 1.0))
 ```
 

AGENTS_TEMPLATES.md

@@ -150,7 +150,7 @@ fn main() {
         .with_fitness(MinimizeDistance { target: 0.5, precision: 1e-5 })
         .with_fitness_ordering(FitnessOrdering::Minimize)
         .with_select(SelectTournament::new(0.5, 0.02, 4))
-        .with_crossover(CrossoverMultiPoint::new(0.7, 0.8, 3, false))
+        .with_crossover(CrossoverUniform::new(0.7, 0.8))
         .with_mutate(MutateMultiGene::new(10, 1.0))
         .call()
         .unwrap();
@@ -349,7 +349,7 @@ fn main() {
         .with_max_stale_generations(1000)
         .with_fitness(HyperparamFitness { precision: 1e-5 })
         .with_select(SelectTournament::new(0.5, 0.02, 4))
-        .with_crossover(CrossoverSingleGene::new(0.7, 0.8))
+        .with_crossover(CrossoverUniform::new(0.7, 0.8))
         .with_mutate(MutateSingleGene::new(0.2))
         .call()
         .unwrap();

CLAUDE.md

@@ -51,7 +51,7 @@ parameter reference, copy-paste templates, and troubleshooting.
 Located in `src/genotype/`, defines the representation of solutions:
 - `BinaryGenotype`: Boolean alleles (Vec<bool>)
 - `ListGenotype<T>`: List of values from a fixed set of alleles
-- `UniqueGenotype<T>`: Permutation of unique values
+- `UniqueGenotype<T>`: Positional permutation (swap-only mutation, default T = usize)
 - `RangeGenotype<T>`: Numeric values within a range (default f32)
 - `MultiListGenotype<T>`: Per-gene allele lists
 - `MultiUniqueGenotype<T>`: Multiple unique sets

src/genotype/unique.rs

@@ -13,11 +13,11 @@ use std::hash::Hash;
 
 pub type DefaultAllele = usize;
 
-/// Genes are a vector of unique values, taken from the allele_list using clone(), each value occurs
-/// exactly once. The genes_size is derived to be the same as allele_list length. On random
-/// initialization, the allele_list are shuffled to form the genes. Each pair of genes has an equal
-/// probability of mutating. If a pair of genes mutates, the values are switched, ensuring the list
-/// of alleles remains unique. Defaults to usize as item.
+/// Genes are a vector of values taken from the allele_list using clone(). The values don't need to
+/// be unique, they are only treated as positionally unique (never changed, only swapped). The
+/// genes_size is derived to be the same as allele_list length. On random initialization, the
+/// allele_list is shuffled to form the genes. Each pair of genes has an equal probability of
+/// mutating. If a pair of genes mutates, the values are swapped. Defaults to usize as item.
 ///
 /// # Panics
 ///

src/lib.rs

@@ -18,6 +18,15 @@
 //! * [Genotype](crate::genotype): Knows how to generate, mutate and crossover chromosomes efficiently
 //! * [Fitness](crate::fitness): knows how to determine the fitness of a chromosome
 //!
+//! **Important**: [FitnessValue](crate::fitness::FitnessValue) is `isize` (not `f64`). This
+//! enables equality checks for staleness detection. For float-based fitness, scale manually:
+//! `Some((score / precision) as FitnessValue)`, or use the
+//! [fitness_value](crate::fitness::fitness_value) helper.
+//!
+//! AI agents: see
+//! [AGENTS.md](https://github.com/basvanwesting/genetic-algorithm/blob/main/AGENTS.md) for
+//! decision matrices, constructor parameter reference, copy-paste templates, and gotchas.
+//!
 //! All multithreading mechanisms are implemented using [rayon::iter] and [std::sync::mpsc].
 //!
 //! ## Quick Usage