Agentic Manual Testing — New Chapter in Agentic Engineering Patterns Guide
Simon Willison's Blog·high signal
Willison published a new chapter on manual testing patterns for coding agents. Core principle: "The defining characteristic of a coding agent is that it can execute the code that it writes." Covers three testing approaches: `python -c` for Python libraries, Playwright for browser automation, and curl for JSON APIs. Establishes the norm that agents should always verify their own code — never assume LLM-generated code works without execution. Directly actionable for any builder using Claude Code,