Featured
Agents as Code: Treating AI Agents as Versioned Artifacts
We version our infrastructure. We version our pipelines. We version our policies. So why are most teams still treating their AI agents like throwaway prompt...
Agentic Workflows: Reimagining Repository Automation with Natural Language
Repository automation works well for deterministic tasks, but struggles with work requiring context understanding and adaptive decision-making. When you need to keep documentation synchronised with...
All Stories
Agents as Code: Treating AI Agents as Versioned Artifacts
We version our infrastructure. We version our pipelines. We version our policies. So why are most teams still treating their AI agents like throwaway prompt snippets floating in someone’s clipboard?...
Agentic Workflows: Reimagining Repository Automation with Natural Language
Repository automation works well for deterministic tasks, but struggles with work requiring context understanding and adaptive decision-making. When you need to keep documentation synchronised with code, triage issues intelligently, or...
MCP Apps: Interactive Dashboards Inside Your AI Conversations
If you’ve been working with AI agents lately, you’ve probably hit a limitation: the agent can describe what it found, but it can’t show you anything interactive. Your agent finishes...
GitHub Copilot's Agentic Memory: Teaching AI to Remember and Learn Your Codebase
One of the biggest challenges with AI coding assistants has been their stateless nature with every interaction starting from scratch, requiring developers to repeatedly explain coding conventions, architectural patterns, and...
Beyond Planning: How GitHub Spec Kit Transforms Ideas into Implementation
In my journey exploring agentic DevOps tools, I’ve previously written about delivering major enhancements using Awesome Copilot and GitHub’s built-in Plan Agent. Each tool has pushed the boundaries of what’s...
