Navigate skill graphs via deterministic random walks. Fuses derivational chains, algebraic structure, color determinism, and bidirectional flow for skill recombination.
Status: ✅ Production Ready
Trit: +1 (PLUS - generative recombination)
Principle: skill_{n+1} = walk(seed_n, graph_n)
Frame: Skills as nodes, concepts as edges, walks as derivations
Random Walk Fusion traverses skill graphs using deterministic random walks to discover novel skill combinations. Each step derives from the previous via seed chaining, producing reproducible concept-blending paths.
seed₀ → skill₀ → concept₀ → seed₁ → skill₁ → concept₁ → ...
| Source Skill | Contribution | Integration |
|---|---|---|
| unworld |
| Derivational chains |
| Walk succession is derivational, not temporal |
| acsets | Algebraic structure | Skills form C-set: functor from schema to Set |
| gay-mcp | Color determinism | Each step gets deterministic (color, trit) |
| world-hopping | Bidirectional flow | Walks are reversible via involution |
# Walk step: derive next position from current state + skill trit
next_seed = (current_seed ⊕ (skill_trit × γ)) × MIX mod 2⁶⁴
next_skill = skills[next_seed mod |skills|]