← All domains

AI & Agent Tooling

Agent runtimes, the protocol layer above them, and retrieval that holds up in production.

How I use it

This is the domain I'm deepest in right now. Not because I've read the papers, but because I've had to make the unglamorous decisions: where the isolation boundary sits, what a skill is allowed to do, when retrieval is worth the complexity, and when a rule beats a model.

Depth

TechnologyDepthWhat I did with it
MCP (Model Context Protocol)ExpertShipped a plugin platform and skill repository on it; built a personal server.
Agent sandbox architectureExpertIsolation model, runtime boundaries, and blast-radius design.
Agentic orchestrationExpertHybrid rule-engine plus model routing for operations workflows.
Agent context / knowledge basePracticedStructured markdown as retrievable, self-contained context.
RAG architecturePracticedChunking strategy, retrieval quality, and governance guardrails.
Multi-agent designPracticedSpecialist routing and escalation paths.
Claude Agent SDKPracticedAgent tooling and skill authoring.

Where I think this goes

The category's real bottleneck isn't model quality any more — it's everything around the model. Isolation, permissions, retrieval quality, evaluation, cost control. Products that treat those as infrastructure concerns to be solved later are the ones that stall at pilot.

Platform & Infrastructure →Data & Analytics →