ai-engineering-curriculum

name	ai-engineering-curriculum
description	Structured AI engineering curriculum — 382 skills + 99 prompts across 20 phases covering ML, deep learning, LLMs, agents, and production systems. Use when learning AI, building AI skills,.

When NOT to Use

• Task is outside your authorization scope
• You need to implement controls (use implementing-* skills)
• Task is about analysis, not action (use analyzing-* skills)
• You don't have access to target systems
• Task requires compliance expertise (consult professionals)
• Task is about defense, not offense (use defensive skills)

Overview

AI Engineering from Scratch: a complete 473-lesson curriculum spanning 20 phases from math foundations to autonomous agent systems. Covers Python, TypeScript, Rust, and Julia. Built for agents and humans who need a structured path through modern AI engineering.

Source: rohitg00/ai-engineering-from-scratch

Install: npx skills add rohitg00/ai-engineering-from-scratch

Capabilities

• Navigate a structured 20-phase AI engineering curriculum
• Identify skill gaps by mapping current knowledge to phases
• Recommend targeted lessons based on learner level
• Run placement quizzes to find starting phase
• Run per-phase comprehension checks

When to Use

Trigger phrases:

• "ai engineering curriculum"

• "User wants to learn AI engineering from scratch"

• "Need to assess someone's AI skill level"

• "Looking for structured learning path in ML/DL/LLMs/Agents"

• User wants to learn AI engineering from scratch

• Need to assess someone's AI skill level

• Looking for structured learning path in ML/DL/LLMs/Agents

• Building AI training programs or onboarding materials

• Searching for specific AI topic coverage

Phase Map

Phase	Topic	Focus
0	Setup	Environment, tools, Python/TS/Rust/Julia config
1	Math	Linear algebra, calculus, probability, statistics
2	ML Fundamentals	Supervised/unsupervised, evaluation, pipelines
3	Deep Learning	Neural networks, backprop, CNNs, RNNs
4	Computer Vision	Image classification, detection, segmentation
5	NLP	Text processing, embeddings, sequence models
6	Speech/Audio	ASR, TTS, audio processing
7	Transformers	Attention, encoder-decoder, positional encoding
8	Generative AI	GANs, VAEs, diffusion models
9	Reinforcement Learning	Policy gradient, Q-learning, PPO
10	LLMs from Scratch	Tokenization, training, scaling laws
11	LLM Engineering	Fine-tuning, RAG, prompt engineering, evals
12	Multimodal AI	Vision-language models, cross-modal reasoning
13	Tools/Protocols	MCP, function calling, tool use patterns
14	Agent Engineering	ReAct, planning, memory, tool orchestration
15	Autonomous Systems	Self-improving agents, reflection, verification
16	Multi-Agent/Swarms	Agent coordination, delegation, consensus
17	Infrastructure/Production	Serving, monitoring, scaling, cost optimization
18	Ethics/Safety/Alignment	RLHF, red teaming, guardrails, interpretability
19	Capstone	End-to-end project combining all phases

Built-in Agents

• Primary agent handles core task execution
• Validator agent checks output quality
• Reporter agent formats and delivers results
• Each agent operates with clear input/output contracts

/find-your-level

Placement quiz that assesses current knowledge across phases and recommends a starting point.

/check-understanding

Per-phase quiz that tests comprehension after completing each phase.

Usage

User: "Where should I start learning AI?"
Agent: Runs /find-your-level placement quiz, maps results to phase map, recommends starting phase

User: "I know Python and basic ML, what next?"
Agent: Maps to Phase 3-4, recommends deep learning and computer vision modules

User: "Quiz me on transformers"
Agent: Runs /check-understanding for Phase 7

How to Use

1. Invoke the skill when relevant domain keywords appear in the request
2. Provide required inputs as specified in the skill definition
3. Review the output for correctness before delivering to the user
4. Combine with related skills for complex multi-step workflows

Verification

After completing this skill, confirm:

• Output meets the defined quality and completeness requirements
• All prerequisites are verified and documented
• Error handling covers edge cases
• Results are accurate and actionable

Process

1. Analyze the task requirements
2. Apply domain expertise
3. Verify output quality

Anti-Rationalization

Rationalization	Reality
"I will add monitoring later"	Without monitoring, you cannot detect failures. Add it from day one.
"One model is enough"	Different tasks need different models. Route intelligently.
"Premature optimization"	Infrastructure decisions are hard to change later. Design for scale early.