#agentic-workflow

GitHub agentic workflows and Renovate

I’ve been a big fan of Renovate for a couple of years already. Renovate scans your repositories, detects outdated package versions, and…

ReubenOrg — The Risk Guardrail, or Your Agents Are Too Leaky

A few weeks back one of my reubies (my agents) needed an email address out of a credentials file. It printed every string in the file, then…

Surge AI — 100페이지 회사 규정을 AI 에이전트는 정말 지키고 있을까? HANDBOOK.md 벤치마크 파헤치기

1. 한눈에 보는 핵심 요약 (Executive Summary)

Why Your Hinges Need Guardrails

If I were a dictionary,¹ my word of 2026 would be “guardrails.”

Modernizing sourcing at scale: How intelligent agents are changing procurement

Picture this: It’s late in the week, an urgent sourcing decision is due, and sitting in your inbox are three separate supplier price sheets…

Agentic AI with LangChain — Part 3: Tool Calling in LangChain

This is the third installment in the Agentic AI with LangChain series. In Part 2, we built an agentic RAG workflow orchestrated with…

Where Agentic AI Cost Actually Goes

Why cheaper model calls can still create more expensive agentic workflows

The Prompt Is Not the Problem. The Workspace Is.

People think they need better prompts. What they really need is a system that stops making them start over.

Implementing RAPTOR: How to Build Hierarchical RAG for Complex Document Processing

Raptor Workflow

Coding Agents Are Stochastic — So Increase Reliability by Designing Protocol

TL;DR:

A CRM monitoring agent

My mother runs a home electrification consulting business in Toronto, Canada: heat pumps, solar, insulation, the whole retrofit thing…

Nought to sixty: your first proven Agentic AI workflow

The fastest route to agentic AI is deliberately unglamorous: baseline, assist, prove, then automate. In that order.

AI Agent Anti-Patterns (Part 5): The Illusion of Control

📚 Series Navigation: Part 1: Agent Design Anti-Patterns | Part 2: Tooling & Scale Anti-Patterns | Part 3: Knowledge and Document…

Let the Agent Complain: A Feedback Loop for Better AI Tools

Why agent complaints can become a feedback loop for better AI testing tools.

【菜鳥 PM 學習日記】EP 12｜不只是叫 AI 幫忙寫文件：我們如何重新設計 PM 的工作流程

PM AI Workflow 終於試行約一個月啦！分享我們最後建立的「文件撰寫五步驟」，以及這些日子的觀察和心得。

We Built a Slack-Native AI Agent for the Full Dev Lifecycle

How Axon takes a team from Jira ticket to merged PR.

Make Claude Talk Straight

Claude is smart. But out of the box, its answers can be a lot. Long paragraphs. Jargon. A wall of reasoning with no clear “so what do I do…

Tavsiyeden Eyleme: Perakende Sektöründe Proje Tamamlama Odaklı Yapay Zeka Ajanı İş Akışları

Perakende sektörü, basit LLM tabanlı sohbet araçlarının ötesine geçerek, minimum insan müdahalesiyle karmaşık ve çok aşamalı süreçleri…

Looping vs Prompting: When Should You Let the Model Think, and When Should You Control the Process?

Large Language Models (LLMs) have fundamentally changed how we build software. Unlike traditional applications, where every decision is…

[Tutorial] Building a Multi-Agent Travel Planner with Google ADK

Most travel apps optimize for popularity and convenience, often directing users toward overcrowded, carbon-intensive destinations.

Pattern Matching in an AI World: Why the Future Belongs to Entrepreneurs Who See Workflow Patterns…

In the age of AI, raw intelligence or coding prowess alone won’t win. The real differentiator is pattern recognition the ability to spot…

My Experience at the NVIDIA Workshop in Hyderabad (India Agentic AI Hackathon)

On 27th June, I was RSVP’d for the in person NVIDIA workshop held at Hyderabad, India as a part of the “India Agentic AI Open Hackathon”…

The Agentic Enterprise Needs a Control Plane, Not Another Copilot

For the last two years, enterprise AI has been trapped in a strangely narrow question: Which model should we use?

Governing LLMs in the Enterprise with AI Model Hub — A DevTalk

How do you scale LLM usage without losing control of cost, access, and governance?

Engineering Agentic Guardrails: A Blueprint for Secure Autonomous AI Architecture

Most corporate AI safety frameworks are built for static Large Language Models (LLMs) — systems whose risk profile ends when a text…

The Death of Token Economics: Why “Cost Per Outcome” is the New Return on Investment (ROI)

From Portfolio Builder to ODPR Recipes: Making Data Product SDK Workflows Flexible

The Data Product SDK started with useful commands. Validate a catalog. Build a graph. Render a portfolio. Sync artifacts. Explain outputs…

SuperSpecs: The Discipline Framework That Stops Your AI Agent From Going Rogue

How spec-driven planning, enforced TDD, and a wiki that never forgets finally tame the chaos of AI-assisted software development

Beyond Single Agents: Architecting Reliable Multi-Agent Systems

The first agent deployment is usually a proof of concept dressed up as a production system.

What Is an AI Harness, and How Do You Build One?

The model decides what to do. The harness stores memory, runs tools, enforces rules, and keeps the agent working.

You’re Probably Overpaying for the Tokens That Find Your Vulnerabilities

The most expensive, most sophisticated security model on our benchmark finished tenth. A model that costs a dollar to run beat it. Here’s…

The Framework I Built for Defining AI Micro Agents — and What It Taught Me

The last thing I did at Cubyts was define a framework for building micro agents — and it turned out to be one of the more interesting…

The Power of Neutral Friction: Why the ‘Grill Me’ Skill is a Game-Changer for AI Coding

In the rapidly evolving landscape of AI-assisted software engineering, we routinely fall into the “specs-to-code” cycle: a frustrating…

Install By Prompt: Your Quickstart Is For Agents

The old developer quickstart assumed one reader: a human at a terminal.

Basic Agentic Coding with Grill-Me skill

Since I started writing code with AI agents, I’ve mostly been having them handle repetitive tasks. While I wouldn’t call my codebase folder…

ODPR: A Lightweight Workflow Recipe Specification for Data Product Delivery

Data product work is becoming more structured.

Three Terminals, One Firewall: Getting AI Coding Agents Into a Locked-Down Looker Project

A practical look at running Claude Code, Codex, and Gemini CLI as a team, and the plain text trick that let them work where git could not…

Step-by-Step: Building a High-Cost Therapy Anomaly Detection Agent with UiPath and Snowflake Cortex

Co-authored with Masaki Kumamoto, UiPath

Day 8 — Release Automation: Semantic Versioning From Commit Messages

When you push a commit with `feat:` or `fix:` in the message, the workflow automatically bumps the version, generates a changelog, commits…

How Autyn Cut Loan Processing from 2 Hours to 2 Minutes at 98% Accuracy

Reconstructing borrower income from a stack of unstructured tax returns, paystubs, and bank statements is the most error-prone task in a…

From Prompt to Action: Building Agentic AI That Actually Gets Things Done

The future of AI isn’t about answering questions. It’s about taking action.

OpenCode References: AI Kod Ajanları İçin Yeni Bir Context Katmanı

Artık mesele sadece daha iyi model seçmek değil; doğru repo, doğru doküman ve doğru örneği ajanın çalışma alanına bağlamak.

Agentic AI Explained: How Autonomous Agents Work, Why It Matters, and What I Learned Building One

Spoiler: It’s not a chatbot. It’s a digital worker with a brain, a to-do list, and a conscience.

This Is Bigger Than Advertising

Stop Building GPTs. Start Building Loops.

Prompt, Skill, Subagent ve Flow: AI Ajan Akışlarını Basitçe Anlamak

Yapay zeka dünyası her geçen gün yeni bir jargonla büyüyor — agent, agentic flow, skill, subagent. Bu kelimelerin etrafında öyle bir…

From RPA to Agentic Workflows: Building Resilient AI Architecture for 2026

Enterprise automation has officially entered its most critical evolutionary phase. Over the past decade, organizations poured billions into…

Your First AI Agent Was Practice

It shipped on a Friday. The demo had gone perfectly. Your team had walked the investors through the workflow three times, the agent called…

Scoping AI Coding Assistants: Why Convention Fails and How Enforcement Works

The problem with handing Claude the entire codebase, and what actually works instead.

用 Vibe Coding 打造專屬的 LLM 知識庫！Spicomellus 的 Why & How

最近我燒掉了 Copilot Pro+ 每月額度的一半，做出一個讓我自己每天都在用的東西：一個專門給 LLM 使用、而且能讓我親眼看到 LLM 腦袋裡到底存了什麼的知識庫。我將它命名為 Spicomellus。

Claude Opus 4.8 Prompting Guide: 9 Copy-Paste Prompts for Real Work

Claude Code & Claude AI — Temperature is locked, thinking is adaptive-only and off by default, and effort has five levels. Here are the…