What It Looks Like from Inside the Stack
A first-person account from the AI running inside the environment described in Parts 1 and 2.
This is Part 3 of three. Part 1 covers the architecture. Part 2 walks through a realistic working day.
AI Agents & the Agentic Web¶
What Happens When Your Agent Needs Knowledge From Five Teams?
Part 5 of the KCP series. Previous: How Do You Tell an Agent "This Data Cannot Leave the Building"?
What a 10× Workday Actually Looks Like
A realistic Tuesday with real output numbers, eight tasks, and the parts nobody talks about.
This is Part 2 of three. Part 1 covers the architecture. Part 3 is written by the model running inside it.
Four Layers: How I Built an AI Development Environment That Partly Runs Itself
A technical walkthrough of the Synthesis + Claude Code + Mímir + Klaw stack — what each layer does, how they connect, and why the architecture matters.
This is Part 1 of three. Part 2 walks through a realistic day using this stack. Part 3 is written by the model running inside it.
Who Let the Agent In?
Part 7 of the KCP series. Previous: The Agent Read the Whole Spec. It Didn't Need To.
Add knowledge.yaml to Your Project in Five Minutes
A practical walkthrough of the KCP adoption gradient — from the minimum viable manifest to a full knowledge graph. No theory. Just the steps.
Your AI Has One Layer. It Needs Four.
The debate is "RAG or knowledge graphs?" The answer is neither — and both. Most teams pick one retrieval approach and stop. The interesting question is which layer they are missing, and what blind spot that creates.
KCP and MCP: One Protocol for Structure, One for Retrieval
The previous post
introduced KCP and why llms.txt does not scale to production agent deployments. This post covers
what happens when you connect a knowledge.yaml manifest to a live MCP server — and why the
combination changes how agents behave.
The Date the AI Invented
The agent answered the ROI metrics question with zero tool calls. It reported the indexing speed, the search latency, the file count, the retrieval time improvement, the test count. All correct. Every number accurate.
Then it said the metrics were validated on February 19, 2026.
The actual date was February 17.
We Gave the AI Better Documentation. It Got Slower.
We had 15 skill files documenting every Synthesis CLI command — syntax, options, example invocations, expected output. We wrote them carefully. We loaded them into the agent's context. We assumed the agent would use them.
Then we ran a benchmark.
The CLI condition was the worst-performing integration in the entire test. Worse than no integration at all.