skills.rest
DuqueOM
DuqueOMCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

debug-ml-inference

Diagnose latency and incorrect predictions in production ML inference services.

5|Updated Nov 18, 2025
Search Tags:#debugging#production#fastapi#observability#diagnostics#latency#ml-inference
One-click install
npx skills add https://github.com/DuqueOM/ML-MLOps-Portfolio --skill debug-ml-inference
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: debug-ml-inference
Source: https://github.com/DuqueOM/ML-MLOps-Portfolio/tree/main/.devin/skills/debug-ml-inference
Command: npx skills add https://github.com/DuqueOM/ML-MLOps-Portfolio --skill debug-ml-inference

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Debug ML inference issues in production services such as latency spikes, incorrect predictions, and event loop blocking.

Core Features & Use Cases

  • Systematic diagnostic workflow for production ML inference endpoints (latency, accuracy, timeouts) with evidence collection.
  • Anti-pattern checks, runbooks, and remediation planning to quickly identify root causes and verify fixes.
  • Use Case: When a FastAPI inference service shows slow responses, this Skill guides you through diagnosis, commands, and rollback or improvement steps.

Quick Start

Diagnose why a specified ML inference service is slow or producing incorrect predictions and propose a concrete remediation plan.

Frequently Asked Questions about debug-ml-inference

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

FAQPage Schema
How do I debug ML inference latency spikes in a FastAPI service?▼

To debug ML inference latency, this Skill applies anti-pattern checks (D-01..D-32) to identify root causes like event loop blocking. It captures command outputs and logs to propose deterministic fixes for your FastAPI service.

What is the best way to diagnose incorrect predictions in production ML models?▼

Diagnosing incorrect predictions involves a systematic workflow that checks for inference anti-patterns and gathers evidence. This Skill analyzes service behavior to identify root causes and generates remediation runbooks for production ML models.

How do I find out why my ML inference endpoint is timing out?▼

Finding why an ML inference endpoint times out requires checking for event loop blocking and latency spikes. This Skill applies diagnostic checks, gathers artifacts, and proposes concrete rollback or improvement steps to resolve timeouts.

Can I generate an incident investigation runbook for production ML inference issues?▼

Yes, you can generate an incident investigation runbook for ML inference issues. This Skill captures command outputs, logs, and artifacts during diagnosis to support remediation handoffs and document root cause analysis.

Does this diagnostic workflow work without external observability dependencies?▼

Yes, the diagnostic workflow operates without external dependencies. It independently applies internal anti-pattern checks to analyze service behavior and gather evidence directly from your ML inference environment.

Related Skills

Explore More Software Engineering Skills→
Nir-BhayNir-Bhay
19

ml-api-endpoint

Deploy machine learning models as FastAPI inference endpoints with Dockerfile support.

Community
Advanced
karlmdaviskarlmdavis
1

getting-pr-artifacts

Enumerate GitHub PR workflow artifacts and generate gh download commands.

Community
Intermediate
praxstackpraxstack
1

backend-principle-eng-python-ml-pro-max

Plan and build Python AI/ML backends with data quality and reproducibility.

Community
Advanced
zcacereszcaceres
7

quality-perf-review

Analyze full-stack codebases to identify and remediate performance bottlenecks.

Community
Advanced