Skip to content

scuptio/churcuring

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Churcuring

┌──────────────┐
│   -> { } │.│ │
│ § ->  Σ  [#] │
│   ->  λ  │:│ │
└──────────────┘
AI  -->  Churcuring  -->  MuduDB

Churcuring is an AI-native computation and software construction framework that converts natural language into * formal specifications*, compiles them into hierarchical state machines, and generates executable logic and database access code in a structured, verifiable, and infinitely scalable manner.

The name Churcuring honors the Church–Turing tradition, reflecting the system’s goal of bridging formal computation, symbolic structure, and modern large language models.


Key Features

  • Natural language → formal, executable specifications
  • UML activity diagrams and ER models as compilable specs
  • Hierarchical, infinitely decomposable state machines
  • LLM-generated state transition functions
  • Explicit, controlled database access via MuduDB
  • Formal verification of state machine correctness
  • Clear separation of logic, state, and side effects

Core Pipeline

Natural Language
      ↓
Structured / Formal Specification
      ↓
State Machine (Σ)
      ↓
Transition Functions (λ) + Database Access ([#])

Churcuring treats software as structured state space, not ad-hoc code.


Architecture Overview

1. Specification Layer {}

  • UML Activity Diagrams
  • Entity–Relationship (ER) Diagrams
  • Fully structured and normalized
  • Can be authored manually or generated by LLMs
  • Acts as the single source of truth

2. State Machine Layer Σ

  • Specifications compile directly into state machine code
  • Each state can be:
    • A terminal state
    • A pure transition function
    • A nested sub-state machine
  • Supports unlimited hierarchical refinement
  • Designed for formal verification (safety, liveness, correctness)

3. Function & Data Layer λ [#]

  • Terminal state transitions may be described in natural language
  • LLMs generate:
    • Transition function code
    • Database read/write procedures
  • All side effects are explicit and isolated
  • Persistent state is managed via MuduDB

Symbol Legend

Symbol Meaning
§ AI prompt / natural language input
{} Structured specification
Σ State machine
λ Transition function / Mudu procedure
[#] Database

Typical Workflows

Natural Language → State Machine

Text → Spec → Compiled State Machine

State Refinement

  • Activity diagrams refine existing state machines
  • Any state can be expanded into a sub-state machine

Natural Language → Code

Terminal State Description
        ↓
LLM-generated Transition Function
        ↓
Executable Code

Formal Verification

Each state machine produced by Churcuring is suitable for:

  • Model checking
  • Reachability analysis
  • Safety and liveness proofs

Verification is a first-class design goal, not an afterthought.


Vision

Churcuring proposes a new AI-centric computation architecture where:

  • AI generates structure, not just text
  • Structure compiles into reliable software
  • Systems remain auditable, scalable, and verifiable

It unifies natural language, formal methods, state machines, and databases into a single coherent pipeline.


Project Status

Churcuring is an open research and engineering framework.
Interfaces, compilers, and runtime backends are under active development.


About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors