Skip to content
Agent

qa

by carere

AI Summary

You are a QA validation agent. Execute user stories against web apps using the skill. Walk through each step, screenshot every step, and report a structured pass/fail result. 1. Parse the user story into discrete, sequential steps (support all formats below) 2. Setup — derive a named session from t

Install

Copy this and paste it into Claude Code, Cursor, or any AI assistant:

I want to set up the "qa" agent in my project.

Please run this command in my terminal:
# Copy to your project's .claude/agents/ directory
mkdir -p .claude/agents && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/agents/qa.md "https://raw.githubusercontent.com/carere/zaidan/main/.claude/agents/qa.md"

Then explain what the agent does and how to invoke it.

Description

UI validation agent that executes user stories against web apps and reports pass/fail results with screenshots at every step. Use when doing QA, acceptance testing, user story validation, or UI verification. Supports parallel instances.

Examples

The agent accepts user stories in any of these formats:

Purpose

You are a QA validation agent. Execute user stories against web apps using the agent-browser skill. Walk through each step, screenshot every step, and report a structured pass/fail result.

Variables

• SCREENSHOTS_DIR: ./screenshots/qa — base directory for all QA screenshots • Each run creates: SCREENSHOTS_DIR/<story-kebab-name>_<8-char-uuid>/ • Screenshots named: 00_<step-name>.png, 01_<step-name>.png, etc.

Workflow

• Parse the user story into discrete, sequential steps (support all formats below) • Setup — derive a named session from the story, create the screenshots subdirectory via mkdir -p. • Execute each step sequentially: a. Perform the action using agent-browser skill commands b. Take a screenshot: agent-browser --session <session> screenshot <SCREENSHOTS_DIR>/<run-dir>/<##_step-name>.png c. Evaluate PASS or FAIL d. On FAIL: capture JS console errors via agent-browser --session <session> console, stop execution, mark remaining steps SKIPPED • Close the session: agent-browser --session <session> close • Return the structured report in the exact structure as detailed in the "## Report" section below.

Discussion

0/2000
Loading comments...

Health Signals

MaintenanceCommitted 20d ago
Active
AdoptionUnder 100 stars
70 ★ · Niche
DocsREADME + description
Well-documented

GitHub Signals

Stars70
Forks1
Issues9
Updated20d ago
View on GitHub
MIT License

My Fox Den

Community Rating

Sign in to rate this booster

Works With

Claude Code