Glossary

This glossary provides definitions for terms used throughout the Hatchling documentation. For foundational concepts, see other appendices.

---

A

API (Application Programming Interface): Set of rules and protocols for interacting with software components. See: settings_reference.md

Appendix: Supplementary documentation section for foundational or background information.

C

CLI (Command Line Interface): Text-based interface for interacting with Hatchling through terminal commands. See: open_a_terminal.md

Container: Isolated runtime environment created by Docker for running applications.

D

Docker: Containerization platform used to run Hatchling and Ollama in isolated environments.

E

Environment Variable: Configuration setting that can be modified to change application behavior and often set at the system's level (the whole computer).

F

Feature: Distinct capability or function provided by Hatchling or its ecosystem.

G

GPU (Graphics Processing Unit): Hardware component that can accelerate LLM inference when properly configured.

H

Hatch: The official package manager for the Hatch ecosystem, integrated into Hatchling.

Hatch Environment: Isolated workspace containing specific collections of MCP server packages.

Hatch Package: A packaged MCP server that can be installed and managed through Hatch.

Hatch Registry: Centralized repository of available MCP server packages.

Hatch Schemas: JSON schemas for package metadata and validation.

Hatch Validator: Tool for validating packages against schema specifications.

Hatchling: The main CLI application that provides chat interface with LLM and MCP tool integration.

I

Inference: The process of running a model to generate predictions or outputs.

L

LLM (Large Language Model): AI model used for natural language processing and generation, run locally through Ollama.

M

MCP (Model Context Protocol): Protocol that enables language models to securely access external tools and data sources.

MCP Server: Software component that provides tools and capabilities accessible through the MCP protocol.

MCP Tool: Individual function or capability provided by an MCP server.

O

Ollama: Local LLM runtime that provides API access to language models.

P

Package: A distributable unit of software, such as an MCP server or extension.

PlantUML: Tool for creating diagrams in documentation. See: PlantUML

R

Registry: Centralized location for storing and retrieving packages.

S

Schema: Structured definition for validating package metadata.

Server: Software that provides services or tools accessible via a protocol.

Settings: Configuration options for Hatchling and related tools. See: settings.md

T

Terminal: Program for entering CLI commands. See: open_a_terminal.md

Tool Calling: Capability that allows LLMs to invoke external functions and tools during conversations.

Tutorial: Step-by-step guide for users. See: tutorials

U

User: Person interacting with Hatchling or its ecosystem.

V

VRAM: Video memory on GPU cards, important for determining which LLM models can run effectively.