Glossary

Core Concepts

Agent

An AI system that can reason about problems, plan actions, use tools, and iterate until reaching a goal. Agents combine LLM reasoning with tool access.

Example: A ReAct agent solves "What is 10 + 5" by calling a calculator tool.

ReAct

Pattern for agent operations: Reason → Act → Observe → Repeat

The agent thinks about what to do, executes an action (tool call), observes the result, and repeats if needed.

Tool

A function that an agent can call to interact with external systems or perform computations.

Examples: calculator, database query, API call, file system operation

Model Context Protocol (MCP)

A standard protocol for tools and agents to communicate. An MCP server exposes tools; an MCP client (like an agent) discovers and uses those tools.

LLM

Large Language Model. An AI system (like GPT-4, Claude) that processes text and can be used for reasoning and decision-making in agents.

Architecture Concepts

Event Sourcing

A pattern where state changes are recorded as immutable events. The current state is reconstructed by replaying past events.

Benefits: Full history, auditability, recovery, temporal queries

Microservices

An architecture where a system is composed of multiple small, independent services that communicate over APIs.

Examples: Jenkins service, Gitea service, email service

MCP Server

A service that implements the MCP protocol and exposes tools/capabilities to clients (agents).

Message Queue

A system for asynchronous communication between services. Services send messages; other services process them later.

Example: AWS SQS

• Jenkins publishes: "Build #42 completed"
• Watcher agent subscribes and reacts to the event

Tools & Technologies

Docusaurus

A static site generator for building documentation sites. Used for the Fellowship exercises documentation.

Docker/Docker Compose

Docker containerizes applications. Docker Compose runs multiple containers together.

• Container: Isolated environment with code + dependencies
• Image: Template for a container
• docker-compose.yml: Configuration file defining all services

Jenkins

CI/CD automation server. Manages build pipelines.

Common Operations: Create job, trigger build, check status, view logs

Gitea

Git hosting platform (like GitHub). Manages repositories and version control.

Common Operations: Create repo, push code, open PR, manage releases

code-server

Browser-based VS Code editor. Allows coding from any machine via web browser.

MailHog

Local email testing service. Captures all emails sent by the system for testing.

Engineering Patterns

Polling

Repeatedly checking a system for changes.

Example: Watcher agent polls Jenkins every 30 seconds: "Is the build done?"

Webhook

A callback mechanism where a service notifies others when something happens.

Example: Gitea webhook triggers Jenkins build when code is pushed

Tool Chaining

Using multiple tools in sequence to accomplish a complex task.

Example: Agent reads file → parses data → calculates result → stores result

Graceful Degradation

System continues functioning at reduced capacity when something fails.

Example: If email service is down, log the emails instead of failing

Development Concepts

Sandbox / Test Environment

An isolated environment for testing without affecting production.

In Exercises: Your EC2 instance is a sandbox for learning

Type Safety (TypeScript)

The compiler checks data types before running code, catching errors early.

Benefit: Fewer runtime errors

Virtual Environment (Python)

An isolated Python environment for a project. Each project can have different dependency versions.

Command: python -m venv venv

Test Coverage

Percentage of code tested by automated tests.

Goal: High coverage = fewer bugs

DevOps / Cloud Concepts

EC2 Instance

Virtual machine on AWS. Your learning environment runs on an EC2 instance.

S3 Bucket

Cloud storage on AWS. Files are stored in buckets.

CloudFront

Content Delivery Network (CDN). Caches content at locations worldwide for fast access.

Example: Documentation is cached at edge locations near users

IAM

Identity & Access Management. Defines who can do what in AWS.

SSM Parameter

A secure way to store configuration values (like API keys, URLs) in AWS.

Workflow / Process Terms

CI/CD

• CI: Continuous Integration - code changes trigger automated tests/builds
• CD: Continuous Deployment - passes builds are automatically deployed

GitHub Actions

GitHub's workflow automation system. Runs jobs (tests, builds, deployments) automatically.

Artifact

Generated output from a build (binary, documentation, code package).

Example: exercises-v1.2.3.tar.gz is an artifact

Communication / Collaboration

Synchronous

Communication where both parties are present at the same time.

Example: Real-time chat, phone call

Asynchronous

Communication where parties respond on their own time.

Example: Email, message queue, webhook

API

Application Programming Interface. A contract for how software communicates.

Example: Jenkins API lets you trigger builds programmatically

REST API

Common API style using HTTP methods (GET, POST, PUT, DELETE).

Common Abbreviations

Term	Full Name	Meaning
MCP	Model Context Protocol	Standard for agents to use tools
LLM	Large Language Model	AI system like GPT-4
ReAct	Reasoning + Acting	Agent pattern
CI/CD	Continuous Integration/Deployment	Automated testing and deployment
API	Application Programming Interface	Communication contract
JSON	JavaScript Object Notation	Data format
REST	Representational State Transfer	API architectural style
SDK	Software Development Kit	Library for using a service
IDK	I Don't Know	Honestly, ask for help!

Ready to Learn More?

• See Architecture Overview for system design
• Check Development Guide for implementation details
• Start with Exercise 1!