Generate reproducible skill combinations via deterministic random walks on skill graphs.
npx skills add https://github.com/plurigrid/asi --skill random-walk-fusion
Please help me install this Agent Skill. Skill: random-walk-fusion Source: https://github.com/plurigrid/asi/tree/main/skills/random-walk-fusion Command: npx skills add https://github.com/plurigrid/asi --skill random-walk-fusion
Navigate skill graphs with deterministic random walks to discover novel skill combinations by fusing derivational chains, algebraic structure, color determinism, and bidirectional flow for recombination.
Start with a genesis seed and run a forward walk to observe a reproducible path, then use the path to recombine concepts into a new skill.
High-intent search queries and answers about installing and using this skill.
Deterministic random walks navigate skill graphs to fuse derivational chains and discover novel skill combinations. Start with a genesis seed to generate a reproducible walk path, then use that path to recombine concepts into a new synthesized skill.
Deterministic walks produce identical paths from the same seed, enabling reproducible skill discoveries. This means you can reliably trace the same derivational relationships and recombination logic multiple times without randomness.
Yes, walks are reversible to explore alternative recombinations. Bidirectional flow lets you navigate forward through derivational chains and backward to discover inverse skill fusions and complementary concept paths.
GF(3) conservation and ACSet-based graph schemas provide the algebraic foundation. Color and trit attribution create deterministic branching, while the Skill/Concept/Walk topology enables forward, backward, triadic, and hop explorations for reproducible recombination.
Run a forward walk using your genesis seed as the deterministic anchor. The seed-based generation applies GF(3) conservation rules across the skill graph, producing a consistent path that you can then use to fuse skills into new ideas.
Use deterministic walks when you need reproducible, systematic discovery of novel skill fusions at scale. This approach excels when skill graphs have rich derivational relationships and algebraic structure that manual exploration would miss.
plurigridGenerate deterministic wide-gamut LCH colors from seeds using SplitMix64 and GF(3) trits in Julia/Jupyter notebooks via Gay.jl.
equinorDiscover Fusion skills via Fusion MCP with GitHub-backed catalog fallback.
subimagesecBuild Cypher queries for SubImage Neo4j relationship investigations.
plurigridConstruct directed bridges and perform hops across ordered locales with GF(3) conservation.
