Glossary - CybMASDE
This glossary provides short definitions of terms used across the CybMASDE documentation. For detailed information and examples, consult the linked wiki pages.
General Terms
API
Application Programming Interface, a contract that lets components (or applications) exchange data and commands. See also the CLI and the Library API: cli_reference, lib_reference.
Dependency
A library or package required by a project module. See the installation procedure: installation.
Virtual Environment
Isolated Python environment (virtualenv/venv) used to manage dependencies without affecting the system Python. See manual installation: installation.
Workflow
A sequence of automated (CI/CD) or operational steps. Example automation in GitHub Actions: faq (CI section).
MkDocs
A static site generator used for the documentation: mkdocs.
GitHub Actions
CI/CD service to automate tests and deployments. See FAQ and contribution guide: faq, contributing.
CybMASDE-Specific Concepts
MAMAD
Core methodology of CybMASDE: Modeling, Analyzing, Monitoring/Adapting, Deploying. Overview and context: architecture.
MOISE+
Organizational formalism used to constrain and explain multi‑agent behavior. Referenced in architecture and training docs: architecture.
MARL (Multi‑Agent Reinforcement Learning)
Family of reinforcement learning algorithms for multi‑agent systems (examples: MAPPO, MADDPG, QMIX, ROMA). See Training section: architecture.
World Model
Learned model (latent dynamics, representations) that approximates the environment to accelerate or stabilize training. See library reference: lib_reference.
Joint Policy
A shared or coordinated strategy among agents resulting from multi‑agent training. Examples and export options: cli_reference.
Auto‑TEMM / TEMM
Analysis and explainability methods used to infer and explain organizational structures and behaviors. See Library/API reference: lib_reference.
Transferring
Deployment and synchronization of trained policies to real or remote environments (via REST API). See usage and FAQ: getting-started, faq.
Refining
Post‑deployment improvement iterations (retraining or adjusting configuration based on analysis). See Next Steps: getting-started.
Interfaces & Tools
CLI ( cybmasde )
Command-line interface to orchestrate the pipeline (init, validate, run, deploy, export...). Full CLI reference: cli_reference. The CLI entrypoint is documented throughout the CLI reference pages.
GUI (Angular)
Angular-based graphical interface for creating/editing projects and visualizing metrics. GUI guide: gui_reference.
Python API
Programmatic usage via the Python library (example: CybMASDEProject ). See the Getting Started and Library API pages: getting-started, lib_reference.
REST API / IPC
Communication modes between backend ↔ frontend or backend ↔ external environments:
- REST API for network exchanges (remote deployment).
- IPC for local integration (Electron). Architecture details: architecture.
Algorithms & Models Mentioned
- VAE, RNN... Architectures for learned dynamic representations (world models). See: lib_reference.
- MAPPO, MADDPG, QMIX, ROMA... Example MARL algorithms mentioned in the docs: architecture.
Configuration Files & Schemas
project_configuration.json
Main project configuration file describing project steps, paths, and parameters. Usage reference: getting-started.
label_manager.py
Optional script to manage labels/custom project hooks (referenced by the GUI docs): gui_reference.