AI Debug Tool: Evidence-Based Diagnosis for ML Experiments

AI Debug Tool: Evidence-Based Diagnosis for ML Experiments

The AI Debug Tool is a specialized diagnostic skill designed to methodically investigate and diagnose issues within machine learning experiments. It emphasizes an evidence-based approach to debugging, systematically probing systems before hypothesizing causes for issues like training failures, divergence, out-of-memory errors, hanging processes, and abnormal performance metrics.

What this skill does

• Probes the system to gather initial data, including process lists to check for multiple instances or defunct processes, kernel or system events, GPU statistics, log scrollbacks, and checkpoint states, all before forming a hypothesis.
• Once preliminary evidence is gathered, it formulates a hypothesis regarding the potential cause of the issue.
• Performs a ‘smoke test’ to assess the hypothesis' validity before making a conclusion about the root cause.
• Supports a disciplined debugging process following the probe → hypothesis → smoke → controls → claim protocol.
• Automatically intervenes when an AI assistant speculates a cause without evidence, routing through the AI Debug Tool instead.

Who it is for

• Machine Learning Developers and Engineers working with AI coding agents like Claude Code, Cursor, and Codex.
• Data Science Teams responsible for managing and optimizing ML experiments.

Use cases

• When machine learning experiments consistently fail or show unusual behavior such as increasing loss, zero GPU utilization, or unexpected system utilization patterns.
• During troubleshooting requests where clarifications on crash logs, performance issues, or process statuses are needed.

Technical details

• Utilizes tools compatible with AI coding agents including Claude Code, Cursor, and Codex.
• Incorporates process state checks, system event logging, and hardware status evaluations.
• Ensures controlled and systematic debugging workflow, improving the accuracy of diagnosing ML experiment issues.

Source & Licence

This package is built on open-source work published by fcakyon (fcakyon/labrat) and distributed under MIT. The original licence text and copyright notice are included in your download.

Personal and commercial use, modification and redistribution are permitted, provided the original copyright and licence notice are retained.

Your purchase covers curation, licence verification, packaging, documentation and instant delivery. It does not grant exclusive rights to the underlying open-source code, which remains available under its original licence.

Delivery & Support

• Delivery: instant — a secure download link is emailed to you as soon as payment is confirmed.
• Format: ZIP archive containing the skill files, documentation and the original licence.
• Support: support@mcpcart.com — we aim to reply within 2 business days.
• Updates: updates are included only where stated on this page.

Refunds

This is a digital product delivered immediately after purchase. By completing your order you request immediate delivery and acknowledge that, once the download has been accessed, the statutory right to cancel no longer applies to the extent permitted by law. Refund requests are handled in accordance with our published Refund Policy.

Claude, Codex, Gemini and Cursor are trademarks of their respective owners. MCP Cart is an independent marketplace and is not affiliated with, endorsed by, or sponsored by any of them. Compatibility references describe interoperability only.