Enclave
A theoretical architecture for reactive, persistent narrative systems that separates human authorship, authoritative world state, actor memory and cognition, provenance, event validation, and probabilistic inference.
View projectAI systems infrastructure
Infrastructure for durable AI systems.
Laughing Skull develops persistent memory, semantic state, repository intelligence, runtime coordination, and supporting tools for AI systems that need to operate beyond a single prompt, model, or session.
Approach
State should outlive the model.
Long-lived AI systems need more than larger context windows. Durable state, explicit ownership, provenance, temporal change, scope boundaries, and deterministic verification need to exist independently of any single inference call.
Models interpret and propose. Deterministic systems retain authority over identity, state, provenance, validation, and persistence.
Deterministic stateObservation + contextBounded model judgmentDeterministic verificationDeterministic persistence
Major work
Persistence, context, and agent operation.
Reliquary
A persistent semantic storage and runtime system that separates source evidence, scoped memories, entities, relationships, derived observations, temporal state, provenance, ownership, and retrieval—designed so durable knowledge remains inspectable and provider-independent.
View projectWarlock
A persistent local work environment for agents and software tools that coordinates repository state, semantic analysis, graph visualization, documents, jobs, services, and diagnostics.
View projectToolchain
Deterministic systems beneath the agent layer.
Lexicon + Arcana
Lexicon publishes deterministic polyglot repository facts. Arcana turns those facts into queryable graph state for symbols, calls, dependencies, impact analysis, paths, call chains, unresolved references, and architecture.
View projectPitlord
Evaluates repository-owned architecture and policy rules against source and Arcana graph state, producing deterministic findings, evidence, baselines, diffs, and diagnostics.
View projectArchivist
Repository documentation integrity and governance for schemas, links, indexes, codemaps, maintenance rules, and documentation-aware checks.
View projectOpen source
Built in the open where it makes sense.
Laughing Skull develops public infrastructure and tooling alongside private experimental work. Public repositories are intended to be inspectable technical evidence: working systems, architecture, tests, and implementation history rather than marketing-only demonstrations.
About
Independent systems work from British Columbia, Canada.
Laughing Skull is an independent software and AI systems company focused on the infrastructure required for AI systems that persist, accumulate state, operate across tools, and remain intelligible over time.
People. First. Always.
Contact