Building Realistic Environments for AI Agent Evaluation
The tools, data, permissions, and starting conditions that bring real workflows into controlled, repeatable agent evaluations.
Latest Articles
Articles on AI evaluation workflows, benchmarking methods, applied AI systems, and Quantiles product updates.
AI Agent Documentation: A Guide to llms.txt, AGENTS.md, SKILL.md, and MCP
An overview of the documentation, instructions, and protocols that help coding agents understand developer products, complete workflows, and work more reliably.
Using Token Data to Improve the AI Agent Experience
Token usage as a signal for agent behavior, workflow efficiency, reliability, and friction in developer products.
Does your SKILL.md help coding agents?
How to evaluate whether agents find and use your skill, how it affects their behavior, and whether it helps or harms the use of your product.
Evaluating AI Agent Tool Use
Understand how coding agents discover, select, and use tools, and how those behaviors affect successful use of your product.
Measuring Agent Recovery
How to verify recovery from failures, measure recovery effort, and evaluate how reliably AI agents resolve problems and continue tasks.