Best of Medium — August 2025

1. 1
   Article

   Medium·41w

   Authentication Explained: When to Use Basic, Bearer, OAuth2, JWT & SSO

   Authorization controls what users can do after authentication through three main models: RBAC assigns permissions to roles, ABAC uses attributes and context for fine-grained control, and ACL attaches permissions to individual resources. Real applications like GitHub and Stripe often combine these models. OAuth2 enables delegated authorization without sharing credentials, while JWTs and bearer tokens carry user identity and permissions across systems. The key is choosing the right combination of models and mechanisms based on your application's complexity and security requirements.

   937

   25
2. 2
   Article

   Medium·42w

   Mastering n8n: Step-by-Step Beginner’s Guide

   A comprehensive beginner's guide to n8n, a visual automation platform that connects apps and automates workflows without coding. Covers setup options (cloud vs self-hosted), understanding the workflow canvas, five types of nodes (trigger, action, logic, code, AI agent), testing and debugging strategies, and building modular systems. Introduces agentic workflows that can reason through context and adapt to changing conditions, with practical examples for social media automation, sales research, data analysis, and personal AI assistants.

   458

   15
3. 3
   Article

   Medium·42w

   The Best Software I’ve Ever Used Was Written by One Person

   Solo developers create superior software because they maintain complete creative control and focus on solving specific problems rather than pleasing committees or investors. One-person projects feel more coherent and deliberate, with every feature earning its place through ruthless prioritization. While these tools risk disappearing when their creators move on, they offer an intimate user experience and resist the bloat that comes from corporate growth strategies. The author argues these projects represent acts of resistance against extraction-focused software design, providing focused tools that prioritize user needs over monetization.

   140

   10
4. 4
   Article

   Medium·41w

   Build Your First Agent in 5 Days

   A practical 5-day guide for building AI agents from scratch, covering essential tools like OpenAI GPTs, n8n for automation, CrewAI for multi-agent systems, Cursor for AI-powered coding, and Streamlit for user interfaces. The guide provides a structured approach with a reusable recipe framework breaking agents into brain, tools, orchestration, interface, and hosting components, plus a complete Python code example for a basic agent implementation.

   110

   2
5. 5
   Article

   Medium·39w

   The Bill Gates Approach: How to Answer “What Are Your Salary Expectations?” in Tech Interviews

   A strategic approach to answering salary expectation questions in tech interviews, inspired by Bill Gates' value-focused mindset. The framework emphasizes researching market rates, leading with value proposition rather than numbers, and considering total compensation packages. The key insight is to frame salary discussions around the impact and problems you'll solve rather than treating it as a confrontational negotiation.

   106

   12
6. 6
   Article

   Medium·43w

   Stop Writing Try/Catch Like This in Laravel

   Identifies five common anti-patterns in Laravel exception handling that lead to production issues: blind catching of all exceptions, empty catch blocks, logging without context, transaction abuse, and controller bloat. Provides specific fixes including targeted exception catching, contextual logging, proper transaction scoping, and centralized exception handling through Laravel's global exception handler.

   96

   7
7. 7
   Article

   Medium·40w

   5 Agent Workflows You Need to Master (And Exactly How to Use Them)

   Five structured AI agent workflows are presented to replace ad-hoc prompting: prompt chaining breaks complex tasks into sequential steps, routing directs queries to appropriate models based on complexity, parallelization runs independent tasks simultaneously, orchestrator-workers use a planning model to coordinate specialized workers, and evaluator-optimizer creates feedback loops for quality improvement. Each workflow includes Python code examples and addresses specific use cases like code generation, content creation, and data analysis to achieve more consistent and production-ready results.

   81

   1
8. 8
   Article

   Medium·42w

   7 PHP Operators Every Laravel Pro Should Master

   Modern PHP operators can significantly improve code readability and safety. The null-safe operator (?->) eliminates nested null checks, null coalescing assignment (??=) simplifies default value setting, and the spread operator (...) streamlines array merging. The match expression provides a cleaner alternative to switch statements with strict comparison, while the spaceship operator (<=>) simplifies sorting logic. Understanding the difference between null coalescing (??) and Elvis (?:) operators prevents logic bugs, and chaining null coalescing operators creates clean fallback chains.

   71

   6
9. 9
   Article

   Medium·43w

   FrankenPHP vs PHP-FPM (Part 3): CPU, Memory, and the Hidden Cost of Doing Nothing

   A comprehensive performance comparison between FrankenPHP and PHP-FPM focusing on CPU and memory usage patterns during both active load and idle states. The analysis reveals that FrankenPHP with worker mode delivers 3x better throughput and 60% lower latency than PHP-FPM, but consumes more memory even when idle (85MB vs 25MB). PHP-FPM excels for low-traffic applications due to minimal resource usage when inactive, while FrankenPHP becomes more cost-effective for high-traffic scenarios due to its persistent application state and reduced per-request overhead.

   41

   7
10. 10
    Article

    Medium·43w

    Google Gemini 2.5 Deep Think : The best ever AI is here

    Google released Gemini 2.5 Deep Think, an AI model that prioritizes reasoning over speed by taking more time to think through problems. Available to Ultra subscribers, it achieved bronze-level performance at the International Math Olympiad and leads benchmarks in coding and mathematical reasoning. Unlike traditional chatbots that respond instantly, Deep Think uses parallel reasoning and reinforcement learning to test ideas and revise answers, making it ideal for complex problems in mathematics, coding, and design. The model is more cautious and less prone to hallucination but may refuse some harmless requests.

    36

    3
11. 11
    Article

    Medium·39w

    GPT-5 System Prompt Leaked : 7 Prompt Engineering Tricks to learn

    Analysis of a leaked GPT-5 system prompt reveals seven key prompt engineering techniques including identity locking to prevent prompt injection, knowledge anchoring for temporal context, multimodal toggles for routing, personality injection for behavioral control, content safety as first-class instructions, self-denial of hidden mechanisms to prevent conspiracy theories, and dynamic retrieval gates for up-to-date information. The techniques demonstrate advanced strategies for building robust AI systems through careful prompt design rather than fine-tuning.

    29

    1
12. 12
    Article

    Medium·41w

    What’s new in Flutter 3.35

    Flutter 3.35 introduces significant productivity improvements including stable stateful hot reload for web development, experimental Widget Previews for isolated component testing, and enhanced accessibility features. The release includes 1108 commits from 168 contributors, with new Material and Cupertino components, improved Impeller performance, and better multi-platform support. Key additions include the Dart and Flutter MCP Server for AI-assisted development, Android sensitive content protection, and various UI component enhancements. The update also brings breaking changes including component theme normalization, Radio widget redesign, and increased minimum Android SDK requirements.

    19

    1