audit-usability

Audit CLI, API, and web interfaces for ISO/IEC 25010 usability issues.

11|1|Updated Jul 25, 2025

Search Tags:#static analysis #accessibility #error messages #usability audit #iso 25010 #api ergonomics #cli ux

Ready to run

npx skills add https://github.com/tomzx/agents --skill audit-usability-tomzx

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: audit-usability
Source: https://github.com/tomzx/agents/tree/main/skills/audit-usability
Command: npx skills add https://github.com/tomzx/agents --skill audit-usability-tomzx

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Usability defects like missing help text, cryptic error messages, unguarded destructive commands, and accessibility gaps are easy to overlook during development and painful for users afterward. This Skill performs a read-only static audit of a repository's CLI, API, and web surfaces against the ISO/IEC 25010 usability characteristic and produces a severity-classified findings report. ## Core Features & Use Cases - Multi-surface scanning: Detects CLI frameworks (argparse, click, typer, cobra, yargs), HTTP API routes, and web templates/components, then audits each surface that exists. - Five usability dimensions: Checks recognizability (help text, OpenAPI docs), learnability (examples, quickstarts, naming consistency), operability (--version, exit codes, defaults), user error protection (actionable error messages, confirmations on destructive actions), and web accessibility (alt text, form labels, keyboard reachability). - Evidence-based findings: Classifies findings by severity (critical to low) with evidence levels, executing decisive checks in /tmp scratch scripts without modifying the repository. - Use Case: Before a public API release, run the audit with the api surface flag to find undocumented endpoints, inconsistent error responses, and missing input validation messages. ## Quick Start Ask the agent to run a usability audit on this repository's CLI surface and report any missing help text, poor error messages, or unguarded destructive commands.

Frequently Asked Questions about audit-usability

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How are usability findings severity classified?▼

Critical covers unguarded destructive operations and commands with no help at all. High covers non-actionable error messages and missing input validation. Medium covers missing --version or inconsistent flag naming, and Low covers missing examples or minor naming inconsistencies.