Cognitive Walkthrough

When you need to figure out why users keep getting stuck on a specific flow, this walks you through a task step by step from a novice user's perspective. It asks four questions at each action: will they know what to do, can they find the control, does it look like it does what they need, and will they get feedback. You break down something like "create account and add to wishlist" into atomic steps, then systematically identify where confusion happens. The output is concrete, with severity ratings and a predicted success rate for first-time users. It's more surgical than running general heuristics, best when you're debugging a critical journey like checkout or onboarding and need to know exactly where people will bail.

skills.sh installs: 502
GitHub stars: 69
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add mastepanoski/claude-skills --skill cognitive-walkthrough --agent claude-code

Installs into .claude/skills of the current project.

Files

Cognitive Walkthrough

This skill enables AI agents to perform a task-specific usability evaluation using the Cognitive Walkthrough method, a technique that simulates how users (especially novices) think through completing specific tasks in an interface.

Unlike broad heuristic evaluations, Cognitive Walkthrough provides deep analysis of particular user journeys, identifying where users get stuck, confused, or make errors.

Use this skill when you need granular, task-focused insights into learnability and ease of first use.

Combine with "Nielsen Heuristics" for general usability, "Don Norman Principles" for intuitive design, or "WCAG Accessibility" for inclusive access.