Notes on making agent experiences reliable.
Uber Has 5,000 MCP Tools. Can Agents Actually Find the Right One?
How Uber makes thousands of tools available to agents, and what an incident investigation can teach us about finding tools, using them, and retrieving useful evidence.
Agent experience (AX): Comparing four social media API providers
We tested four social media API providers across 204 complex research tests. See how their agent experience (AX), costs and completion times compare.
An example of why Parallel beat Exa in our AX eval
A shortened preview was only the start. Parallel gave the agent a concrete way to inspect the saved response and keep researching.
Ten search providers. Can agents actually use their interfaces?
Hundreds of research trials across ten search providers revealed instruction conflicts and response-handling friction. We examine agent experience alongside completion, cost and time.
Designing CLIs for Better Agent Experiences
Design a CLI as an agent interface: task flows, structured output, dry runs, and feedback loops.
