debug-plan

A browser debug plan extracts URLs and classifies user interactions—navigate, fill, click, verify—into structured test steps. Provide your debugging scenario as a prompt; the Skill parses it into a ready-to-execute plan with URL targets and step-by-step actions confirmed before execution.

Yes. This Skill integrates into Phase 1 of a debug workflow, parsing prompts to generate browser test plans. It enforces URL resolution, interaction taxonomy, viewport settings, and capture configurations, outputting a HITL-confirmable test structure compatible with Chrome DevTools automation.

Describe the debugging scenario in natural language: the URL or page path to test, the interactions you want to perform (navigate, fill form fields, click elements, verify results), and any specific conditions. The Skill extracts this context and infers the environment to build a concrete, executable plan.

Yes. The Skill classifies interactions across a defined taxonomy—navigate, fill, click, verify, and others—allowing multi-step test plans within a single prompt. Each step is structured as JSON schema-compliant test steps with viewport and capture settings.

The Skill outputs a structured plan with URL targets and ordered steps. Before execution, it prompts for human-in-the-loop confirmation, allowing you to review, adjust, or approve the plan before it runs in the debug workflow.

Yes. The Skill accepts natural-language prompts describing what to debug. It handles URL extraction, environment inference, and interaction parsing automatically, returning a ready-to-confirm test plan without manual schema or configuration.