ICE Studio is an Integrated Cognitive Environment
for system-level software development.
It is not an IDE.
It is not an AI tool.
It is not a plugin-based editor.
It defines a new category of system.
👉 Read the official ICE Studio documentation:
https://francescomaiomascio.github.io/ICE-Studio-Docs/
This site is the canonical reference for understanding ICE Studio: its architecture, runtime model, cognitive principles, and design philosophy.
The ICE Studio engine and core implementation are intentionally not open source.
This repository hosts documentation only.
Private source repository:
https://github.com/francescomaiomascio/ICE-Studio
ICE Studio is the reference implementation of ICE Engine —
a cognitive runtime designed to orchestrate intelligence across:
- code
- knowledge
- agents
- runtime
- interfaces
ICE Studio does not extend an IDE.
It replaces the IDE abstraction entirely.
- an IDE
- a chatbot
- a plugin ecosystem
- an LLM wrapper
- a browser engine (e.g. Chromium)
- a game engine for cognition
- an operating runtime for knowledge-driven systems
ICE Engine is the underlying runtime that powers ICE Studio.
ICE Engine does not “think”.
ICE Engine orchestrates thinking.
It coordinates:
- cognitive agents
- local and remote LLM models
- structured knowledge layers
- lifecycle phases and transitions
- system-wide observability
- human–system interaction
- modular interfaces and protocols
The engine exists independently from any single project or UI.
ICE Studio is built around cognitive phases, not screens.
A deterministic bootstrap phase responsible for:
- environment verification
- capability discovery
- runtime topology (local / remote)
- system identity initialization
- minimal cognitive context creation
Preboot is not a setup wizard.
It is a cognitive bootstrap phase.
The Dashboard represents the global cognitive environment.
It provides:
- system-wide awareness
- access to Global Knowledge
- orchestration across projects
- runtime diagnostics and logs
Dashboard ≠ workspace
Dashboard is the system observing itself.
A project is a living cognitive entity.
Each project has:
- its own knowledge
- history and timeline
- dedicated agents
- contextual objectives
Projects plug into a living system — they do not recreate it.
ICE Studio enforces strict separation between knowledge layers.
System-level knowledge about architecture, capabilities, protocols, and runtime state.
Used by:
- System Agent
- orchestrators
- decision engines
Does not contain user or project data.
Represents long-term patterns, preferences, and behavior.
Built through observation, not prompts.
Local, private, and isolated.
Project-specific context:
- code
- documents
- decisions
- objectives
- evolution timeline
This separation prevents contamination and enables targeted reasoning.
The backend is the cognitive engine.
It is responsible for:
- agent orchestration
- knowledge routing
- lifecycle control
- logging, audit, and tracing
- model integration
This is not a simple API server.
Agents are not chatbots —
they are specialized cognitive processes with single responsibilities.
The GUI is a surface, not the core.
It:
- renders state
- collects input
- mounts and unmounts phases
It does not:
- store knowledge
- make decisions
- perform reasoning
Preboot, Dashboard, and Project phases never coexist in the same runtime.
ICE Studio enforces centralized, event-based logging.
Direct file writes are forbidden.
All logs must:
- emit events via
ice_studio.logging.api - be routed through the logging runtime
- remain observable and auditable
ICE Studio supports:
- agent plugins
- UI plugins
- knowledge plugins
- distributed runtimes
Internal protocols (ICEP / ICENet) enable:
- runtime discovery
- remote coordination
- cognitive coherence across nodes
If you believe in the vision behind ICE Studio and want to support its development:
Your support helps sustain long-term research, system design, and the evolution of ICE Studio.
ICE Studio is not an IDE.
It is what comes after.