eaccmk

Well-Known MCP Concepts

Concept	Definition	Example Use Case	Reference Link
Tool Invocation	AI agent calls external APIs or tools in a standardized way.	An agent fetches invoice details by calling an accounting API via MCP.	[Merge.dev][web:162]
Session Memory	Persistence of conversational context across steps or platforms.	User pauses a product return on mobile, resumes later on web with all context preserved.	[ActivePieces][web:161]
Cross-Platform Context Sharing	Transfer of history, context, and tasks across devices/apps.	U

eaccmk

Term	Definition	Example Use	Reference Link
MCP (Model Context Protocol)	Protocol enabling AI agents to share structured context and communicate tool usage.	MCP lets AI agents coordinate by sharing what tools and memory they use for complex tasks.	Auth0 Blog
Agent	Autonomous AI component that acts independently to complete tasks using reasoning and learning.	An agent autonomously schedules meetings and sends invites without human input.	Perplexity Blog
A2A (Agent-to-Agent)	Communication protocol for AI agents to interact securely and collaboratively.	A2A allows a chatbot agent to pass a user query to a specialized knowledge agent for an accurate answer.	Google Blog
ACP (Agent Communication Protocol)	IBM-driven standard for structured messaging between

eaccmk

Acronym / Keyword	Definition	Example Use	Reference Link
AGI	Artificial General Intelligence: AI with human-level cognitive abilities across diverse tasks.	Researchers aim to develop AGI capable of performing any intellectual task a human can.	Shelf AI Glossary
API	Application Programming Interface: Protocols for building and interacting with software applications.	Developers use APIs to integrate AI features into their apps seamlessly.	NVIDIA AI Workbench
Attention Mechanism	Neural network component enabling focus on important parts of input data.	Transformers use attention mechanisms to understand relationships between words in a sentence.	Hugging Face Glossary
Benchmark	Standardized tests to evaluate AI model performance on given tasks.	The team used ImageNet as a benchmar

eaccmk

#	Building Block Concept	Short Description	Reference Link
1	Foundation Models	Large AI neural networks trained on massive raw data, enabling broad understanding and generation tasks.	NVIDIA Blog - Foundation Models
2	Machine Learning Basics	Core ML concepts like supervised, unsupervised learning foundational to AI applications.	Microsoft Learn AI Fundamentals
3	Natural Language Processing & LLMs	Techniques for understanding and generating human language using models like GPT and LLaMA.	NVIDIA Blog - Open Reasoning Models
4	AI Model Training & Evaluation	Methods for training models, tuning hyperparameters, and measuring model accuracy and robustness.	OpenAI & NVIDIA Open Models

eaccmk

{
     "data":
     {
         "products":
         [
             {
                 "desc": "A fixed term investment for businesses at an agreed rate of interest.",
                 "name": "90 Day Business Term Deposit"
             },
             {

eaccmk

// author  : https://github.com/eaccmk
// created : sep-2021

//Advanced jsx in fitbit

const user = {
  firstName: "Fitbit",
  lastName: "Buddy",
};

eaccmk

// author  : https://github.com/eaccmk
// created : sep-2021

function mySettings(props) {
  return (
    <Page>
      <ImagePicker
        title="Image Selector"
        description="This feature gets image from phone to fitbit"
        label="Click here to select an image"

eaccmk

// author  : https://github.com/eaccmk
// created : sep-2021

function mySettings(props) {
  return (
    <Page>
      <AdditiveList
        title="Items with Autocomplete feature"
        settingsKey="autocomplete-list"
        maxItems="3"

eaccmk

// author  : https://github.com/eaccmk
// created : sep-2021

function mySettings(props) {
  return (
    <Page>
      <Select
        label={`Click for Selection`}
        settingsKey="selection"
        options={[

eaccmk

// author  : https://github.com/eaccmk
// created : sep-2021

function mySettings(props) {
  return (
    <Page>
      <TextInput
        title="Click me to type"
        label="Type (one, two ...)"
        placeholder="Type here"