What is the benefit of using ATR over running tests directly?

While direct commands only show that a failure occurred, ATR initiates an investigation to explain why it failed and provides a list of recommended fixes.

How does ATR Analyze identify the cause of a failure?

It uses an agentic workflow to parse error outputs, scan mentioned source files, and execute diagnostic commands to pinpoint the exact logic or configuration error.

Which AI models does ATR use for analysis?

ATR supports Gemini-based models, offering a 'flash' tier for rapid feedback and a 'pro' tier for deep, complex troubleshooting.

Can I provide extra context to the AI for specific failures?

Yes, you can use the --context flag to inform the agent about recent changes, such as 'Added new auth module', to narrow down the investigation.

Does this skill support both Python and JavaScript?

Yes, ATR includes automatic environment detection for Python (virtualenvs, Poetry) and Node.js (nvm, package.json engines).

Agentic Test Runner (ATR) Analyze

Name: Agentic Test Runner (ATR) Analyze
Author: imyousuf

byimyousuf

•

Security & Testing

Automates the investigation and debugging of test failures and command errors using AI-powered root cause analysis.

The atr-analyze skill leverages the Agentic Test Runner (ATR) to streamline the debugging process by automatically investigating command failures within Claude Code. When a test or build script fails, this skill triggers an AI agent that proactively examines error logs, reads relevant source files, and performs diagnostic actions to identify the root cause. By providing actionable recommendations and maintaining a clean coding context, it eliminates the manual trial-and-error usually required to resolve complex integration issues, environment-specific bugs, and regression failures.

Key Features

01Intelligent environment detection for Python (venv/Poetry) and Node.js (nvm) projects

02Support for dual model tiers (Flash for speed, Pro for deep multi-file analysis)

03Automated AI-powered failure investigation and root cause analysis

041 GitHub stars

05Clean context management that keeps diagnostic noise out of your primary chat

06Context-aware diagnostics that read source files and error logs automatically

Use Cases

01Debugging cryptic CI/CD or build failures without manual log diving

02Automating the analysis of shell command failures during complex refactoring

03Investigating regression tests where the failure point is not immediately obvious

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add imyousuf/agentic-test-runner atr-analyze

For use in Claude.ai and ChatGPT

Download Skill