skills.rest
Mrlyk
MrlykCOMMUNITY
Explore all skills by this author →
GitHub

sh-qa

Run superharness qa CLI to generate qa-report.md and qa-issues.json.

4|Updated Apr 5, 2026
Search Tags:#automation#qa#cli#quality-assurance#qa-report#superharness
One-click install
npx skills add https://github.com/Mrlyk/superharness --skill sh-qa
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sh-qa
Source: https://github.com/Mrlyk/superharness/tree/main/skills/sh-qa
Command: npx skills add https://github.com/Mrlyk/superharness --skill sh-qa

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill provides a simple, AI-invokable wrapper around the superharness qa CLI so users do not need to run commands manually or inspect raw files to understand external QA outcomes. It centralizes QA execution and reporting into the AI session, reducing friction when validating completed tasks.

Core Features & Use Cases

  • Invoke external QA services: Runs the configured external QA checks for a specific task using the superharness qa command.
  • Generate human- and machine-readable reports: Ensures qa-report.md (human-readable) and qa-issues.json (machine-readable) are written to the task directory for review and automation.
  • Summarize results and recommend next steps: Provides totals for checks run, passed and failed counts, lists issues by severity, and suggests follow-up actions such as running the fix workflow or finalizing the branch.
  • Use Case: After completing a task, run this skill to execute external QA, get a concise summary of problems, and decide whether to run automated fixes or proceed to merge.

Quick Start

Ask the AI to run the superharness qa command for the target task path under .superharness/tasks and summarize the resulting report.

Frequently Asked Questions about sh-qa

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

FAQPage Schema
How do I run external QA and summarize findings for a completed task?▼

To run external QA, invoke the superharness qa CLI for a task stored under .superharness/tasks. This generates qa-report.md and qa-issues.json in the directory, summarizing checks, passed/failed counts, and issues by severity.

What is automated QA reporting and how does it work with the superharness CLI?▼

Automated QA reporting uses the superharness CLI to execute configured external checks on a specified task. It produces a human-readable qa-report.md and machine-readable qa-issues.json, summarizing issues and suggesting follow-up actions.

Can I use this QA automation tool without manually configuring external services?▼

No, you must have the superharness CLI installed and external QA services configured in .superharness/config.yaml. The tool relies on these configurations to trigger automated checks and produce the final reports.

Do I need to inspect raw files manually after running external QA checks?▼

No, the tool centralizes QA execution and reporting into the AI session. It summarizes totals for checks run, passed and failed counts, and lists issues by severity, eliminating the need to inspect raw files manually.

What is the best way to review QA issues and decide whether to run automated fixes?▼

The best way is to review the summarized findings, which list issues by severity and recommend next steps. This allows you to decide whether to run the fix workflow or proceed to finalize the branch.

Why does my QA report generation fail when running the superharness qa command?▼

QA report generation fails if the specified task path is not under .superharness/tasks, or if external QA services are not configured in .superharness/config.yaml. Ensure the superharness CLI is installed and properly configured.

Related Skills

Explore More Software Engineering Skills→
MrlykMrlyk
4

sh-fix

Fix pending QA issues from qa-issues.json using test-driven development.

Community
Advanced
MantisWareMantisWare

code-review

Score branch diffs against guidelines and generate fix prompts.

Community
Advanced
yoshisadayoshisada

kiln-qa-pass

Automate end-to-end QA passes with a four-agent team and GitHub issue reporting.

Community
Advanced
tamayotchitamayotchi

forgecad-render-inspect

Inspect ForgeCAD models with the render inspect command and summarize results.

Community
Advanced