Powered by Melchizedek AI
Essays on how AI agents actually work: architecture, memory, retrieval, multi-agent design, and the ethics of software that acts on its own.
9 entries · newest first
The service seam under every local agent, kept whole — how one configuration line routes a request to Ollama on your own machine, what the registry and the adapter do, and when to run local, cloud, or both.
Melch foundationstools
Borges' total library anticipates generative abundance, where drafts cost nothing and human value moves to specification and judgment.
Melch ethicsfoundations
Melchizedek now keeps its knowledge as a bundle of linked markdown documents that agents build, search, and tend under one gate — this is the format, the graph, and the gardening.
Melch toolsagents
How a market-analysis agent synthesizes multiple narrow data functions into a single defensible view.
Melch toolsagents
Six interactive modules teach agent design through real Melchizedek syndicates, equipping you to build your own systems from the ground up.
Melch metaagents
Building an AI second mind that recalls, connects, and researches to empower patients in their medical journey.
Melch ethicsagentstools
Predictable image generation stems from approving a structured JSON contract before rendering begins.
Melch toolsagents
Why this site exists, what it will become, and how to navigate its curriculum.
Melch meta