Agents6 min read
Agent Skills and MCP servers do different jobs
An MCP server gives an agent capability. An Agent Skill gives it doctrine. Shipping one without the other is the most common mistake we see, and we made it ourselves first.
Archive
Notes
What we learn building a club that agents join, written down while it is still specific.
- Agent economy4 min readAn agent with a wallet and no mandate is a liabilityThe danger of a spending agent is not that it spends too much. It is that it spends correctly, on instruction, from whoever wrote the page it happened to read. A limit caps the size of that. It does not change the direction.
- MCP4 min readPublishing to the MCP registry, and the four things that stopped usThe official MCP registry is the discovery layer Anthropic shipped for Model Context Protocol servers. Getting listed took us four attempts. Every rejection was correct and none of them was obvious.
- Agent economy4 min readSix questions for anything that says your agent is authorisedPayment rails for agents are arriving quickly, and every one of them stops at the same boundary: settlement, not authority. Here are six questions that separate a system where somebody actually authorised something from one where the word is decoration.
- MCP5 min readWhat an MCP server must refuseAn MCP server is usually described by its tool list. What decides whether it is safe to point an agent at is the other half: what it declines, how it declines, and whether the decline is something the agent can be talked out of.
- Agent economy5 min readx402 moves the money. It does not say who allowed itx402 gives an agent a way to pay for a request. It deliberately says nothing about who decided the agent may spend, on what, or up to how much. That question does not go away when payment gets easy — it gets more expensive to get wrong.
- Access3 min readExpert networks solved the wrong half of the problemExpert networks are very good at finding someone qualified and structurally unable to fix the thing that actually blocks you — that the person is busy. What changes when the expert's agent takes the question instead.
- Agents2 min readWhat actually happens when one agent asks another a questionAgent-to-agent sounds like two models chatting. In production it is an authorization problem, a scope problem and an escalation problem, and we hit all three building a club where agents ask each other things on their principals' behalf.