Make Your AI Agent Compare Before It Recommends
Listing alternatives is easy. A useful recommendation explains how each option fits the problem before choosing one.
3 min to read
Blog
Thoughts, tutorials, and insights about web development.
AI Can Write Your Tests. Can You Trust What They Check?
A useful generated test can expose a consequential mistake. Start from the promise, justify the expected result, and inspect which failures the evidence reaches.
7 min to read
I Want an AI Agent That Keeps Going—Until It Shouldn’t
My autohelm skill delegates routine choices inside an authorized task, records consequential decisions, and keeps clear limits on unattended work.
8 min to read
What Happens When a Beginner Gets an AI Agent Before an Expert Does?
A profession's experience barrier may weaken when newcomers learn to deliver useful work with agents before established practitioners change their methods.
5 min to read
Heimdall Uses AI to Give Incident Investigation a Head Start
An AI investigation report can give the on-call engineer current status, useful evidence and a direction to investigate before they open the alert.
3 min to read
I Want an Agent That Challenges My Premise
A coding agent should examine consequential assumptions before implementing them. Useful pushback needs evidence, proportion, and the ability to agree and proceed.
8 min to read
If AI Makes Implementation Cheaper, Why Keep the Same Team Shape?
Separate mockups, specialist implementation and repeated handoffs made sense under particular constraints. Coding agents give teams a reason to reconsider those boundaries.
4 min to read
Stop Fixing Every AI Agent Mistake With Another Rule
Before adding another prohibition, identify the instruction that enabled the failure and check what your proposed correction does to other cases.
8 min to read
What Is Experience Worth When AI Can Write the Code?
When an agent handles implementation, experience still shapes the questions you ask, the mistakes you notice, and the work you can judge.
4 min to read
“The Agent Wrote It” Is Not an Excuse
Claude or Codex can write the implementation. You still own the decision to merge it, the evidence you accept, and what happens when it ships.
7 min to read
Why AI Interfaces Keep Looking the Same
AI can produce a polished interface quickly, but vague direction and screenshot-driven feedback push its designs toward the same familiar patterns.
6 min to read
Decision Models: When AI Should Choose, Rank, or Ask
A decision model turns observations and constraints into a choice, ranking, or deliberate handoff. Learn when that is different from generating text or reasoning through a problem.
8 min to read
MCP, APIs, and A2A: Three Different Ways AI Systems Connect
Understand the boundary between APIs, Model Context Protocol servers, and agent-to-agent communication before adding another integration.
8 min to read
RAG, CAG, or Fine-Tuning: Where Should Your Knowledge Live?
Choose between prompting, context augmentation, retrieval, and fine-tuning by asking what kind of knowledge your AI workflow needs and how quickly it changes.
9 min to read
After an Interruption, Reconstruct What Actually Happened
An interrupted agent needs to inspect files, tool results, jobs, and external state before choosing what to repeat or finish.
7 min to read
AI Can Create More Software and Still Cost Developers Their Jobs
Cheaper software can expand demand while reducing the labor needed for each project. Total growth and an individual developer's prospects are different questions.
5 min to read
AI Removed Some Typing. It Did Not Remove the Mental Load.
Delegating implementation can leave the human work concentrated in review, uncertain decisions, and repeated requests for attention.
3 min to read
An Agent in My Everyday Browser
Chrome Bridge connects an agent to an existing Chrome session, with explicit tab targeting, compact page observations, and concrete focus and permission tradeoffs.
9 min to read
How to Build Beautiful Presentations With AI
Build a talk with a story and a visual identity, then use the HTML presentation kit for navigation and export. Includes an editable eight-slide example.
5 min to read
Insured Vibe Coding: Keeping Agent Builds on Track
Keep an agent build moving through useful phases, meaningful tests and planned checks in the running application, with evidence before completion.
5 min to read
Make Your Coding Agent Lazier About Complexity
A working implementation can still do too much. Give simplicity its own review, keep the requirements fixed, and ask which responsibilities can disappear.
7 min to read
Natural Language Is a Programming Interface. Context Is Part of the Input.
A short follow-up gets its meaning from the task, earlier messages, and corrections that came before it.
3 min to read
Why Agents Lose the Thread After Compaction—and How I Prepare for It
Prepare current project records, verified implementation state, and an explicit handoff before a long agent conversation is compacted.
7 min to read
Let Your Agent Ask You to Step In
An autonomous task sometimes needs a manual action. Give the agent a way to ask, wait for a reply, check the result, and carry on.
3 min to read
Your Coding Agent Is More Than Its Model
Judge a coding agent by the whole system that turns intent into checked work: its tools, context, interface, and communication as well as its model.
7 min to read
I’ve Done 100× More Software Engineering in Two Years With AI Than in My First Six
AI moved my work from repetitive implementation toward plan review, architecture decisions, and learning new systems as I build.
8 min to read
Let a Fresh Agent Review the Decision Maker’s Work
Give a fresh reviewer the requirements, artifacts, and evidence before the builder’s rationale. Keep useful context and make disagreement answerable.
7 min to read
Your AI Agent Joins the Project From Scratch. Document for That.
Make project purpose, boundaries, decisions, and current state discoverable so a fresh coding run can work from the right constraints.
7 min to read
Your Coding Standards May Be Solving Yesterday’s Problem
Agent-led development is a reason to reconsider conventions that police personal taste. Keep the rules that protect behavior, contracts, and future change.
7 min to read
How I Scored 111/120 on the TOEFL iBT
A practical guide to maximizing your TOEFL score through templates, structured practice, and AI-assisted learning—even with limited preparation time.
12 min to read to read
Code Review Agents: Meet the Sissies 💅
A practical, opinionated tour of Claude Code’s multi-agent code review system—and why it quietly raises the bar for frontend teams.
12 min to read
Why You Can't Afford to Miss my AI Fluency Workshop
A hands-on, cross-functional workshop that empowers teams to master AI and AI Fluency for immediate productivity gains and future-proof skills.
12 min to read
5 Feedback Models to Transform Your Team Communication
Explore five practical feedback models to enhance clarity and effectiveness in your team's communication.
10 min to read
Why You Need a Service Layer in Your Frontend Application
Understanding the importance of a service layer in frontend development is a strategic approach to simplify complexities and enhance user experience.
5 min to read
Code Consistency and Quality with Style Guides and Rules
Adapting style guides keeps code clean and consistent, balancing rules with flexibility for tech evolution.
8 min to read
How to Create a Standard Merge Request Description Template
Optimize your collaborative software development with a comprehensive guide to creating standard & efficient merge requests.
10 min to read
I Bet You Don't Know These Sass Features!
Sass has amazing features like loops and if/else that often go unnoticed, although we can make amazing stuff with them in a performant way.
10 min to read
Speed Up Your Programming With VSCode Snippets & Shortcuts
Explore how to configure and leverage VSCode snippets and shortcuts to turbocharge your programming efficiency.
10 min to read
The Five Dysfunctions of a Team
Uncover the keys to overcoming team dysfunction. Learn how to build trust, commitment, and accountability with solutions and examples.
15 min to read
Who Is the 'Ideal Team Player'?
Discover Humility, Hunger, and Smart — key virtues for an ideal team player. Real-life examples and insights from Patrick Lencioni's book.
15 min to read
What Is an ADR (Architectural Decision Record) Document?
Learn about Architectural Decision Records (ADRs), essential documents that capture the context and consequences of significant architectural decisions in software development.
10 min to read
Creating a Git Commit Message Convention
Creating a Clear and structured Git commit messages convention and automatically enforcing it with Commitlint for a clean and informative git history.
10 min to read
How Husky Git Hooks Enhance Code Quality and Fault Prevention?
Boost code quality with Husky git hooks! Pre-commit formatting, commit message standards, and pre-push checks ensure fault-free, confident shipping.
10 min to read
Build Time Data Caching in Next.js
Optimizing Next.js Build: Cut Redundant API Calls with Build Time Data Caching. Learn a Step-by-Step Solution to Boost Next.js Apps' Efficiency!
10 min to read
How To Do Great One-On-Ones as a Manager
Discover the potential of one-on-one sessions for managers and leaders. Learn tips and techniques to create connections with team members.
12 min to read
Front-End Learning Roadmap
Comprehensive senior front-end developer roadmap! Prioritize topics, build a strong foundation, and progress toward mastery.
5 min to read
My Data-Fetching Preference in Next.js
Navigating Next.js data-fetching strategies: A deep dive into SSR, CSR, SSG, ISR, and on-demand revalidation. Make informed choices for optimal performance!
15 min to read
Respecting the LGBT+ Community by Embracing Diversity and Creating Inclusive Spaces
Fostering LGBT+ diversity: Navigating inclusivity with empathy and understanding for a better future.
5 min to read
How to Create an HTTP Proxy in Next.js
Explore the power of HTTP proxies in enhancing authorization security and functionality in a step-by-step guide.
10 min to read
How to Write DRY Redux Code
Mastering Clean and Maintainable Redux Code: A Comprehensive Guide with Redux-Saga Integration in Next.js
10 min to read
Why Do You Need To Rethink About Using Redux
Exploring downsides, complexity, and better alternatives for state management in React projects. Rethink your choices and embrace efficiency.
10 min to read