Resources
Playbooks, guides, and best practices for AI-native E2E testing.
How to Evaluate a Tool That Claims to Be Agent-Native
The label is unverifiable and the mechanism is not. Five questions separate software your coding agent can actually operate from software with agent language on the homepage: can my agent call it, where does the output live, can the agent read the result, what happens on failure, and where is the human.
Where the Human Belongs When Agents Write the Tests
The strongest objection to AI-written tests is not that AI cannot write them. It is that a passing test proves nothing: high coverage, confident test names, and no actual verification underneath. That objection is correct, and the answer is structure rather than trust.
MCP Servers vs Agent Skills: What Each Is For, and Why Tools Need Both
An MCP server gives an agent capability: typed tools it can call at runtime. A skill gives it procedure: an instruction file, invoked as a command, that teaches the agent how to do a job properly. They solve different problems, they fail in different ways alone, and a serious agent integration ships both.
What Is Agent-Native? The Definition, the Disagreement, and What It Means for Quality
Agent-native means software built so an agent is the operator and the human supervises. But six different published definitions are competing right now, and they do not agree on what the term describes. Here is the map, and what the distinction actually changes for testing.
What Is Agent-Native Testing? The Category, the Requirements, and How to Tell
Agent-native testing is end-to-end testing where the coding agent is the operator: it walks the app, writes the tests, runs them, and repairs them, while the human reviews. Here is what that requires architecturally, and how it differs from testing tools that merely have AI in them.
Why Your Tests Belong in Your Repo, Not a Vendor's Database
A test stored as a row in a vendor console cannot be diffed, cannot be reviewed with the code that broke it, does not branch, and cannot be edited by your coding agent. A test stored as a file can do all four. The argument, the honest counter-arguments, and the distinction most comparisons miss: where the artefact lives is a separate question from where it runs.