Conceptual stack

One application. Several explicit layers.

Source files, the assembled application and semantic knowledge are related, but not interchangeable. Keeping that distinction visible is a core part of logiCells.

Model

What is declared?

Project files, explicit concepts and source-level structure.

Application

What is actually assembled?

Resolved packages, metadata, overrides and the effective application structure.

Semantic

What does it mean?

H-Logic, conceptual relations, mappings, constraints and knowledge.

Agent access

Who else can participate?

External agents and tools can work through CLI, Tooling and MCP without redefining application semantics.

Runtime

What can it do?

Policy, confirmation, actions, services and execution over the resolved application.

Why it matters on a client project

Don’t reconstruct intent from a running system.

In most stacks, the model is documentation that fades once the code exists. In logiCells it stays available to the tooling and the runtime. That’s what lets the planner tell you what a change affects, and lets the next developer read what the application means.

See it on a change request
Meaning stays connected
01MODEL: declare
02APPLICATION: assemble
03SEMANTIC: reason
04AGENT ACCESS: participate
05RUNTIME: govern + execute