Skip to main content
Settings
Search
Appearance
Theme Mode
About
Jekyll v3.10.0
Environment Production
Last Build
2026-05-27 23:28 UTC
Current Environment Production
Build Time May 27, 23:28
Jekyll v3.10.0
Build env (JEKYLL_ENV) production
Page Location
Page Info
Layout index
Collection none
Path pages/2/index.html
URL /pages/2/
Theme Skin
SVG Backgrounds
Layer Opacity
0.6
0.04
0.08
page preview image
Creative & Experimental

Generative art, AI fiction, and projects that started as side-quests.

page preview image
Business

Career, freelance, startups, and the economics of a technical practice.

page preview image
AI & Machine Learning

Agentic patterns, Copilot workflows, MCP servers, and ML pipelines.

page preview image
Taming Agent Memory and Context Drift

An agent with no memory reinvents the wheel every time. An agent with bad memory builds on broken foundations. Domain 3 tests your understanding of how to ma...

page preview image
Orchestrating Multi-Agent Workflows on GitHub

One agent is powerful. A coordinated team of agents is transformative. Domain 5 tests whether you can design that team and keep it running.

page preview image
MCP Servers and Agent Tooling in Practice

MCP is the USB-C of AI tooling — a standard connector that lets agents speak to any external system. Here's how to use it on GitHub.

page preview image
Launching the Agentic Codex: A GH-600 Learning Track for the IT-Journey

We built 20 quests, 7 posts, and 8 reference notes covering every skill measured in the GH-600 exam. Here's the map.

page preview image
Evaluating and Tuning Agents with GitHub Signals

You can't improve what you can't measure. Domain 4 tests whether you can define what 'done' means, detect when it isn't, and systematically get better.

page preview image
Embedding Agents in the SDLC: From Tool to Collaborator

Agents aren't just IDE helpers anymore. They can plan, implement, review, and deploy. The question is: where does human ownership end and agent autonomy begin?

page preview image
Agent Guardrails and Responsible Autonomy: The GH-600 View

Autonomy is not a binary. Every agent operates at a level, and every level carries different risk. The question isn't whether to use guardrails — it's where ...