Curated insights on AI agents, A2A workflows, seller strategy, and marketplace design.
Marine Depoorter
Mar 27, 2026 · 8 min read
MCP lets agents use tools. A2A lets agents talk to each other. That's a fundamentally different thing, and it's about to change how we build software.
Fresh notes on buying, selling, and shipping AI agents. The full newsletter flow can come next, but the journal is live now.
Visual: Archive 63-A
As large language models take over the routine, the value of human agency skyrockets. We explore how creators are leveraging autonomous agents to build lean, multi-million dollar micro-studios.
AI Agent Marketplace
Why the next wave of software buying will look less like app stores and more like outcome-driven agent marketplaces with clear delivery, trust, and proof layers.
Sell AI Agents
A practical seller blueprint for solo builders: pick the right delivery model, keep the listing sharp, and use automation only where it reduces friction for the buyer.
A2A Agents
A grounded explanation of A2A agents, where the term is useful, where it is vague, and how buyers should evaluate compatibility instead of relying on buzzwords.
Delivery Models
The operational tradeoffs buyers should evaluate before choosing between hosted access, self-hosted source delivery, Docker images, or downloadable bundles.
Trust Signals
In an agent store, credibility is product infrastructure. Ratings, examples, seller identity, and delivery clarity do more conversion work than slogans ever will.