After the Prompt Series: How Do We Package Everything We Have Built Around the Model?
Agent Plugins let us distribute how we want AI to work. What belongs in the shared core, what depends on the client, and what should a team validate?
Agent Plugins let us distribute how we want AI to work. What belongs in the shared core, what depends on the client, and what should a team validate?
We still need agents, but not every capability deserves its own actor. A practical reflection on responsibility, skills, context boundaries, ALDC Graph, and the cost of agentic complexity.
SkillOpt tests whether small edits to a skill improve an agent. I explore how that process could connect human feedback, validation and ALDC in Business Central development.
Warp turns human feedback into reviewed skill updates. I explore what this could mean for ALDC, AL projects and a partner's engineering knowledge.
Where might HydraFusion fit in AL development? A reflection on adaptive execution, context and evidence, with ALDC as a potential application.
Strategic programming in Business Central, and how to find the line where your own judgement runs out.
Twelve months of framework, and having the right piece from the start still was not enough. When is it ever?
Connecting Microsoft 365 Copilot Cowork to Business Central through native MCP: delegated identity, dynamic action discovery and a clear read-only scope. Includes a guide to reproduce the solution.
Back to the three questions from the earlier posts: is your coding tactical or strategic, how do you see yourself, and where are you — prompts, loop, or graph?
What does each agent receive, what does it return, and how does the system handle our decisions? A sandbox journey through ALDC-GRAPH for Business Central.
We use cookies to improve your experience on this site. By continuing to browse, you agree to our use of cookies.

