From retrieval to agents that do the work
Understand RAG architecture, then explore agents that execute and verify work.
Prerequisites and learning sequenceFeatured paid course
Build agents that execute and verify.
Move from prompts and context to harnesses, loops, and graphs through a portfolio generation and review example.
Developers with basic programming and AI coding experience
Enrollment and payment take place on Inflearn.
- 01PromptDefine the task
- 02ContextAttach rules and files
- 03HarnessEnvironment and checks
- 04LoopRun → review → fix
- 05GraphRoles and handoffs
Harness · Loop · Graph Engineering
9 entries · 5h 10m
Audience, practice and prerequisites- Rating
- 4.8/5
- Reviews
- 24
- Public enrollments
- 238
- Length
- 5h 10m
00CodeBridge
Learn to work with AI —
and to check what it made.
CodeBridge offers courses in AI coding, RAG and agents, and workflow automation. Compare the audience, prerequisites, learning outcomes and practice to choose a course for your current goal.
01Featured course
From using a tool to verifying its work.
An agent design with execution, checks, and handoffs
01 · The problem
Different results with every request
Are you explaining rules and missing checks only after the AI has finished?
02 · The approach
Put the criteria in the environment
Expand prompt-centered agent design into harness, loop, and graph structures to build more controllable systems.
03 · What you can apply
A workflow for the next task, too
Design agent loops with execution and verification / Connect roles and handoffs through graphs
The learning sequence
9 entries · 5h 10m
- Prompts and context
- Working environment and checks
- Execution loops
- Roles and graphs
Entry counts include materials, missions and guidance.
Practice from the public curriculum
Apply the ideas, one stage at a time.
Build a portfolio generation and review workflow with execution, verification, and role handoffs. The course combines concepts with design and practice.
- Who it is for
- For developers who need agents to execute and verify work beyond a single response.
- Before you start
- Basic programming, the ability to run code, and experience with AI coding tools such as Claude Code or Cursor.
- Tools to prepare
- Claude Code, CLAUDE.md, Skills, Hooks, MCP. Tool charges may be separate.
“The explanations are well structured, and the practice was good.”
Excerpt from a public review, translated into English. Read the full feedback, including comments on practice, before choosing.
02All courses
Choose by the problem you need to solve.
AI coding and development
Using Claude Code Properly · Harness Engineering
Connect project rules and verification to the working environment
Familiarity with Git, npm, and project structure, plus experience with Claude Code or similar AI tools.38 entries · 376 minExplore the courseRAG and agent systems
Practical RAG Design · Classic to Agentic RAG
Understand when to use Classic, Graph, and Agentic RAG
Basic RAG and prompting concepts, Python LLM API integration experience, and basic vector database skills.29 entries · 174 minExplore the courseAI CODING
Vibe Code a Chrome Extension with Cursor AI
Turn an idea into a Chrome extension
No coding knowledge required. Basic skills such as creating folders and copying files are enough.12 entries · 61 minExplore the courseAI LEARNING
AI English Growth Routine
Connect reading, shadowing, and AI conversation
Basic English grammar and vocabulary, AI chatbot experience, and basic digital note-taking skills.19 entries · 129 minExplore the courseAI CODING
GitHub Copilot Vibe Coding
Apply Agent and Plan modes to Java and Spring
Basic Java syntax, web concepts, and IDE experience. Basic Spring knowledge is helpful.19 entries · 203 minExplore the courseby goal and experience
4 more courses: agents, Decision AI, web development and content
AI AGENT
Harness · Loop · Graph Engineering
Design agent loops with execution and verification
Basic programming, the ability to run code, and experience with AI coding tools such as Claude Code or Cursor.9 entries · 310 minExplore the courseDecision AI
Practical Decision AI: Laya Models and Human-in-the-Loop
Connect recurring decisions to code
You need to read Python functions and dictionaries and run basic code.8 entries · 108 minExplore the courseAI CODING
AI Website Development · Gemini CLI
Build a website or portfolio with AI
Basic computer skills are needed. Web development experience helps but is not required.10 entries · 107 minExplore the courseAI CONTENT
Introduction to Generative AI Content
Practice AI workflows for writing, images, and video
No coding knowledge required. Basic computer and internet skills are enough.9 entries · 48 minExplore the course03Course finder
Where should I start?
Choose a goal and check the experience needed to see a starting course and why it fits.
Your starting point
Choose your goal.
Check the task and experience needed, then choose a starting course.
Optional next steps and learning sequences
Learning paths
Choose the next step
when you need it.
One course is enough to start. Continue only when you need the next step.
Make AI coding a repeatable workflow
Start with a Java and Spring project, then organize the environment and verification rules.
Prerequisites and learning sequenceFrom repetitive work to your own browser tool
Reduce an Excel task first. Build a browser tool if your web tasks also repeat.
Prerequisites and learning sequence04Free learning
Free course
Beyond using AI coding tools,
build a better workflow.
When AI coding gets stuck, look at the workflow. Six short lessons introduce context, harnesses, loops, and graphs.
Start right away, without signing up.
Curious how AI answers are made? Explore the LLM foundations course.The free course is a separate introduction. For project application, explore the featured course’s practice.
Explore six more free videos
4:42
Open source caught up just a week after Jev's debut
Watch for free
19:00
Harness, Loop & Graph Engineering: five stages of AI agents beyond prompting
Watch for free
11:57
Qwen3.8 vs MuseGlimmer: a hands-on comparison
Watch for free
4:38
Why you shouldn't leave everything to AI: a question raised by Jev
Watch for free
8:15
Cross-platform apps with Qt QML and C++: an introduction to Qt
Watch for free
5:55
Claude Code plugins that save tokens and time
Watch for freeSelected videos are ordered by popularity.
05Blog
Read the engineering behind the tools.
All articlesSoftware Engineering
AlphaProof Code: What a Signal Proof Still Asks Humans to Trust
Read the specification and trust boundary behind 6.5 million lines of Lean proof, and distinguish historical-bug checks from universal vulnerability detection.
AI EngineeringModern Web Guidance: Improving Coding Agent Output Without Changing the Model
LLM training data contains vast amounts of legacy patterns. Modern Web Guidance bridges this gap by providing expert-curated, token-efficient guidelines directly into the agent's context window. Eval results show 27–36 percentage point uplift across multiple agents.
AI CodingHerdr CKD: Worktree Isolation Is Only One Part of Independent Review
Examine the Maker–Checker–Core design, then separate defect detection, integration evidence, and review cost when evaluating a multi-agent workflow.
06Team training, AX and technical consulting
From individual tools
to a team workflow.
Decide what to delegate to AI, which workflows to automate, and how to assess RAG quality. Define training and deliverables around your team’s problems and experience.
- AEngineering teamsProject rules and code verification criteria
- BOperations teamsAutomation candidates and output checks
- CAI systems teamsA design for retrieval, execution and human review
07FAQ
Before you enroll.
How much coding experience do I need?
Basic programming, the ability to run code, and experience with AI coding tools such as Claude Code or Cursor. Beginners can start with free concepts or the workflow automation course.
Where do I check price, payment and access?
Enrollment and payment take place on Inflearn. Check the current listing for price, discounts, access and refund terms.
Are tool subscriptions or APIs extra?
Course fees and tool charges are separate. Check account and pricing requirements for Claude Code, Cursor, Copilot or LLM APIs before enrolling.
Does the English site mean English course audio?
Site language is separate from audio and subtitles. Free videos use Korean audio. Check each Inflearn course for audio, subtitles and translation support.
Where can I ask course questions?
Check the Inflearn listing for course question support and its scope. Use the team inquiry page for corporate training and consulting.
Do I need every course in a path?
Start with one course you need. Further steps are optional, and you can skip topics you already know.
→Start with your next task
Design agent execution and verification together.
Practical AI agent development and automation system design