AI Tutorials: LLMs, AI Agents and MCP
26 posts on ai, newest first.
Every GeekCoding101 post tagged ai is here, newest first. Topics include building AI agents and MCP servers, comparing agentic frameworks, testing Claude Code and other models, and the transformer basics behind them. The posts are hands-on: each one has steps you can run on your own machine, and all are free to read with no sign-up.
Start here
| If you want to… | Read |
|---|---|
| Pick an agentic framework | Agentic Frameworks Compared: LangGraph, CrewAI & More |
| Build an agent yourself | What Is an AI Agent? How to Build One from Scratch in Python |
| Give a model tools over a standard protocol | What Is an MCP Server? How to Build One in Python |
| Lock down an MCP server | MCP Authorization: Protect an MCP Server with OAuth 2.1 |
| Understand how LLMs work | Why is the Transformer Model Called an “AI Revolution”? (start of the Daily AI Insight series) |
AI agents and agentic frameworks
These posts explain what an agent is and compare the frameworks that run agents in practice.
- Agentic Frameworks Compared: LangGraph, CrewAI & More. Explains what an agentic framework is. Compares LangGraph, LlamaIndex, OpenAI Agents SDK, Google ADK and CrewAI, and covers agentic AI coding frameworks such as Claude Code.
- What Is an AI Agent? How to Build One from Scratch in Python. Covers the agent loop, how an agent differs from a chatbot, and a working Python agent with tools built on the Claude API.
- I Ran DeepSeek’s dsh-superagent End to End: A Hands-On Deep Dive. I installed it on a Mac, had it fix a real GitHub issue end to end, and explain its two-layer authorization gate, Mandates and risks.
MCP (Model Context Protocol)
- What Is an MCP Server? How to Build One in Python. Shows how MCP servers expose tools, resources and prompts, and how to build, test and connect one with the Python SDK 2.x.
- MCP Authorization: Protect an MCP Server with OAuth 2.1. Walks through the 401, protected resource metadata and the OAuth 2.1 flow, with a runnable Python SDK server that rejects requests without a valid token.
- **Ghidra MCP: Let AI Help You Reverse Engineer Binaries.** Installs Ghidra 12.1.4 and GhidraMCP 2.0 on an Apple Silicon Mac, then finds zlib’s Adler-32 in a stripped binary over MCP.
Claude Code and model testing
- Claude —dangerously-skip-permissions: What It Does and When to Use It. Explains what the flag really does, why it refuses to run as root, the six Claude Code permission modes, and safer ways to stop the prompts: allow rules, sandbox and auto mode.
- Claude Fable 5.2 Effort Levels, Tested: 26 Runs vs Opus 5.5 and Fable 5.1. Anthropic’s claim is that Fable 5.2 at High matches Opus 5.5 for half the cost, while an independent test found it 4.5x pricier. I ran 26 runs on three models at medium, high and xhigh, with costs taken from transcripts.
Serving models
- **Ray Serve Guide: Autoscaling, FastAPI & Ollama.** Covers deployments, autoscaling, batching and multi-model pipelines, and when to choose Ray Serve over FastAPI or Ollama.
LLM fundamentals: the Daily AI Insight series
These are short explainers from December 2024 for readers new to how large language models work.
- Why is the Transformer Model Called an “AI Revolution”? The first edition of the series.
- 7 Key Insights on the Self-Attention Mechanism in AI Magic. Covers how relevance scores link words, plus real-world uses and limits.
- What Are Parameters? Why Are “Bigger” Models Often “Smarter”? Explains weights, biases and why model size matters.
- What Is Prompt Engineering and How to “Train” AI with a Single Sentence?
- Parameters vs. Inference Speed: Why Is Your Phone’s AI Model ‘Slimmer’ Than GPT-4?
- Discovering the Joy of Tokens: AI’s Language Magic Unveiled. An introduction to tokenization.
- Fine-Tuning Models: Unlocking the Extraordinary Potential of AI. Explains how a pre-trained model is further trained to specialize in a task or domain.
- Ray Serve Guide (above) and OpenAI Unveils o3 and o3 Mini with Stunning ARC-AGI Performance also belong to this series.
Reading “Attention Is All You Need”
- Diving into “Attention is All You Need”: My Transformer Journey Begins!
- Transformers Demystified – Day 2 – Unlocking the Genius of Self-Attention and AI’s Greatest Breakthrough. Works through the paper’s Introduction and Background sections.
- 8 min read
Claude Plans Compared: Pricing, Claude Code and API Costs
Claude plans compared: what Free, Pro, Max, Ultra, Team and Enterprise cost, which include Claude Code, and when a subscription beats pay-as-you-go API billing.
- 7 min read
Linear MCP: Connect Linear to Claude Code Step by Step
Linear MCP is Linear's official hosted MCP server. Add it to Claude Code with one command, sign in with OAuth, then create and update issues by prompt.
- 8 min read
GPT-6.2 Sol vs Astra vs 6.1 Sol: Tested, With Pricing
GPT-6.2 Sol vs Astra vs GPT-6.1 Sol (and Opus 5.5) on the same 9 graded API tasks: accuracy, speed, GPT-6.2 Sol pricing and the real cost per task.
- 24 min read
DeepSeek Harness Hands-On: What dsh Really Runs (and Sends)
A hands-on test of DeepSeek Harness (dsh 0.2.2): install it, count its 299 plugins, write your own tool, and see what each request sends to DeepSeek.
- 6 min read
Funplay MCP: Connect Claude Code to the Godot Editor (Hands-On)
Funplay MCP is an open-source MCP server for game editors. Install the Godot addon, start the server, connect Claude Code and run a first safe request.
- 7 min read
Claude Code Skills: What They Are and How to Write a SKILL.md
Claude Code skills are folders with a SKILL.md that Claude loads on demand or you run as a slash command. See where they live, the frontmatter, and a worked example.
- 9 min read
MCP vs API: What Actually Differs, With a Side-by-Side Demo
MCP vs API: an API is a contract for programmers, MCP is a protocol that lets AI apps discover and call tools. Same notes search built both ways, compared.
- 8 min read
Claude Fable 5.2 Effort Levels, Tested: 26 Runs vs Opus 5.5 and Fable 5.1
Anthropic says Fable 5.2 at High matches Opus 5.5 for half the cost; an independent test found it 4.5x pricier. I ran 26 runs on three models at medium, high and xhigh, with costs from transcripts.
- 9 min read
MCP Authorization: Protect an MCP Server with OAuth 2.1
How MCP authorization works: the 401, protected resource metadata and OAuth 2.1 flow, plus a runnable Python SDK server that rejects requests without a valid token.
- 11 min read
Claude --dangerously-skip-permissions: What It Does and When to Use It
What claude --dangerously-skip-permissions really does, why it refuses to run as root, the six Claude Code permission modes, and safer ways to stop the prompts: allow rules, sandbox, auto mode.
- 27 min read
I Ran DeepSeek's dsh-superagent End to End: A Hands-On Deep Dive
A hands-on test of DeepSeek's dsh-superagent: I installed it on a Mac, had it fix a real GitHub issue end to end, and explain its two-layer authorization gate, Mandates and risks.
- 19 min read
Ghidra 12.1.4 Install (ghidra_12.1.4_PUBLIC) and GhidraMCP
ghidra_12.1.4_PUBLIC is the folder the Ghidra 12.1.4 zip unpacks to. Get it from NSA's GitHub, verify the SHA-256, run it, then add GhidraMCP for AI.
- 9 min read
What Is an AI Agent? How to Build One from Scratch in Python
What makes an AI agent different from a chatbot, the agent loop explained, and a working Python agent with tools built step by step on the Claude API.
- 8 min read
What Is an MCP Server? How to Build One in Python
What the Model Context Protocol is, how MCP servers expose tools, resources and prompts, and how to build, test and connect one with the Python SDK 2.x.
- 20 min read
TypeScript AI Agent Frameworks & Agentic Frameworks Compared
TypeScript AI agent frameworks compared: Vercel AI SDK, Mastra, LangGraph.js and OpenAI Agents SDK, with two runnable examples and a wider comparison.
- 13 min read
Attention Is All You Need, Explained: Intro & Background
A line-by-line walkthrough of the Introduction and Background of "Attention Is All You Need": why RNNs hit a wall and how self-attention replaced them.
- 5 min read
Attention Is All You Need Abstract Explained (Day 1)
The abstract of “Attention Is All You Need” (Vaswani et al., 2017) makes one bold claim. You can build a sequence-to-sequence model out of attention alone, with no recurrence and no convolution, and…
- 5 min read
OpenAI o3 on ARC-AGI: Scores, Cost, Limits
OpenAI's o3 scored 75.7% on ARC-AGI's Semi-Private set and 87.5% at 172x compute. What ARC-AGI tests, what it cost, and why it isn't AGI.
- 9 min read
Ray Serve Guide: Autoscaling, Batching, FastAPI & Ollama
What Ray Serve is, how deployments, autoscaling, batching and model composition work, with code, and when to pick Ray Serve over FastAPI or Ollama.
- 6 min read
Fine-Tuning vs Pre-Training: How LLMs Become Specialists
What fine-tuning is, how it differs from pre-training, the steps to fine-tune a pre-trained model, LoRA vs full fine-tuning, and the risks to watch.
- 8 min read
Tokens in AI: How LLM Tokenization Works
What tokens are, how LLMs split text with word, character and subword (BPE) tokenization, and why tokens shape context windows, cost and speed.
- 6 min read
Parameters vs Inference Speed on Phones
Why phone AI models are smaller than GPT-4: how parameters, memory and hardware set inference speed, and how quantization, pruning and distillation help.
- 8 min read
What Is Prompt Engineering? With Examples
Prompt engineering is writing inputs that steer an AI model to the output you want. Core techniques, before/after examples, limits and when to fine-tune.
- 7 min read
What Are Model Parameters? Why Bigger AI Is Often Smarter
Parameters are the weights and biases a model learns in training. See how parameter count affects LLM performance, memory and cost, and when smaller wins.
- 9 min read
Self-Attention in Transformers, Explained
How self-attention works in Transformers: queries, keys and values, softmax weights, a worked pronoun example, multi-head attention and the O(n²) cost.
- 7 min read
Transformer Model Explained: Why It Was an AI Revolution
How the Transformer and self-attention work, why they replaced RNNs and LSTMs, and their limits. A plain-English guide to the model behind LLMs.
FAQ
Are these AI tutorials free? Yes. Every post on GeekCoding101 is free to read, and there is no sign-up.
I’m new to LLMs. Where should I start? Start with the Daily AI Insight series, beginning with “Why is the Transformer Model Called an ‘AI Revolution’?”. The short posts on self-attention, parameters, tokens and prompt engineering follow it. When you’re ready to build, move on to the AI agent and MCP server tutorials.
What’s the difference between an AI agent, an agentic framework and an MCP server? An AI agent is a model running in a loop: it decides on a step, calls tools and uses the results. An agentic framework is a library that handles that loop, plus state and multi-step orchestration, so you don’t have to write it yourself. An MCP server exposes tools, resources and prompts to a model through the Model Context Protocol, so any MCP-capable client can use them. Each one has its own post above.
Can I follow the tutorials on my own machine? Yes, that’s the goal. The posts include runnable code and real setups. Examples are the Python agent on the Claude API, the MCP server built with the Python SDK 2.x, and GhidraMCP installed on an Apple Silicon Mac.