Blog
Notes on building with multiple AI agents — coordination, shared memory, and keeping context current.
-
29 June 2026
You've become the switchboard between your AI agents
If you run more than one AI coding agent, the problem usually isn't the model. It's that they don't share what they know. Here's why, and what actually helps.
-
29 June 2026
Memory is not the same as a source of truth
AI memory is having a moment, but for teams running multiple agents the harder problem isn't remembering more. It's agreeing on what's true right now.
-
29 June 2026
Giving Claude Code, ChatGPT and Codex the same project context
Most setups make you re-explain your project to every agent and every tool. Here's a cleaner way to keep one shared context they all read from.
-
29 June 2026
Why we built Memeri on MCP
MCP turns 'connect an agent to your tools and data' from a per-tool integration into a shared surface. Here's why we bet the product on it.
-
30 June 2026
Memory vs a shared workspace: what AI coding agents actually need
Everyone's adding memory to AI tools. But if you run multiple coding agents, the gap usually isn't recall, it's a shared place they all work from. Here's the difference.
-
30 June 2026
The coordination tax of running multiple AI coding agents
Every agent you add should make you faster. Past a point it doesn't, because you become the human passing context between them. That's the coordination tax, and how to pay it down.