AI agent skill · by @ClawBio
Genome Match
Score genetic compatibility across all male-female pairings in a Genomebook generation
Archived 2026-10-23 · Source updated 2026-10-04 · License: MIT. Check the original repository for the latest version.
SKILL.md instructions
# 💞 GenomeMatch
## Purpose
Score genetic compatibility between all male-female pairings in a Genomebook
generation. The engine evaluates heterozygosity advantage, disease carrier risk,
and trait complementarity to rank optimal mating pairs for the next generation.
## How It Works
1. **Load genomes** for a target generation from `GENOMEBOOK/DATA/GENOMES/`.
2. **Compute pairwise compatibility** for every M x F combination:
- **Heterozygosity score (40%)**: fraction of loci where offspring would be
heterozygous (genetic diversity advantage).
- **Trait complementarity (40%)**: reward balanced trait combinations and high
average trait values across the pair.
- **Disease risk penalty (20%)**: flag pairs where both parents carry recessive
disease alleles (25% affected offspring risk per flagged condition).
3. **Rank all pairings** by composite score (0.0 to 1.0).
4. **Select non-overlapping mating pairs** via greedy selection from the top of
the ranked list (each individual mates at most once per generation).
## Input
- `GENOMEBOOK/DATA/GENOMES/*.genome.json`
- `GENOMEBOOK/DATA/disease_registry.json`
## Output
- Ranked compatibility table (all M x F pairings)
- Selected mating pairs for the next generation
## CLI Usage
```bash
# Score all pairings for generation 0
python skills/genome-match/genome_match.py
# Score a specific generation
python skills/genome-match/genome_match.py --generation 1
# Demo mode
python skills/genome-match/genome_match.py --demo
# Limit output to top N pairings
python skills/genome-match/genome_match.py --top 10
```
## Output Format
```
Rank Male x Female Score Het Comp Risk Flags
1 einstein-g0 x curie-g0 0.8234 0.650 0.821 0.000 --
2 darwin-g0 x franklin-g0 0.7891 0.600 0.790 0.000 --
...
SELECTED MATING PAIRS (generation 0 -> 1):
Albert Einstein x Marie Curie (compat: 0.8234)
Charles Darwin x Rosalind Franklin (compat: 0.7891)
```