The Agent Skills Directory

[Command Execution via npx]: The skill utilizes the npx utility to execute Playwright commands. This is a common practice for running Node-based developer tools and is used here to power the trace inspection logic.
[Local Web Server for Snapshots]: For viewing and interacting with DOM snapshots, the skill launches a temporary local HTTP server. This allows a headless browser to render the captured HTML content, providing an accurate view of the page state during a test recorded by Playwright.
[File System Interactions]: The skill performs file operations such as extracting ZIP archives and saving screenshots or attachments. These actions are standard for a trace processing tool and are generally directed to the .playwright-cli directory or user-specified paths.
[Browser Automation and Eval]: The snapshot command launches a headless browser instance to execute commands against recorded snapshots. This includes the ability to evaluate JavaScript within the browser context to query DOM elements, which is a key feature for debugging automated tests.
[Data Ingestion Surface]: The skill processes external trace files which may contain data from various web environments. When the agent extracts or queries this data (e.g., via eval), it treats the results as informational content, which is the intended use case for analyzing automated test artifacts.

playwright-trace