retrieving-mlflow-traces

Retrieves MLflow traces via CLI or Python API with flexible filtering options.

1|1|Updated Nov 9, 2025

Search Tags:#debugging #mlops #query #traces #experiment tracking #mlflow

Ready to run

npx skills add https://github.com/Paldom/databricks-apps-streamlit-vibe-coding-starter --skill retrieving-mlflow-traces-paldom

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: retrieving-mlflow-traces
Source: https://github.com/Paldom/databricks-apps-streamlit-vibe-coding-starter/tree/main/.agents/skills/retrieving-mlflow-traces
Command: npx skills add https://github.com/Paldom/databricks-apps-streamlit-vibe-coding-starter --skill retrieving-mlflow-traces-paldom

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps you find and retrieve specific MLflow traces, especially when debugging or analyzing model performance, by allowing you to query them using various criteria.

Core Features & Use Cases

Trace Retrieval: Fetch individual traces by their unique ID.
Trace Searching: Find traces based on experiment ID, run ID, session, user, status, tags, metadata, or execution time.
Debugging: Quickly locate failed traces or traces within a specific session for root cause analysis.
Use Case: When a model deployment shows errors, use this Skill to search for all failed traces associated with that deployment's session ID to pinpoint the issue.

Quick Start

Search for all failed MLflow traces in experiment ID 1.

Frequently Asked Questions about retrieving-mlflow-traces

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

FAQPage Schema

How do I retrieve MLflow traces for debugging model performance issues?▼

Search MLflow traces by filtering on status, execution time, tags, metadata, session, or user via CLI commands or a Python API. This allows you to locate failed runs and pinpoint execution errors efficiently.

Can I filter MLflow traces by execution time and experiment ID?▼

Yes, you can filter MLflow traces by experiment ID and execution time. The retrieval mechanism also supports querying by status, tags, metadata, session, and user for targeted debugging.

What is the best way to locate failed MLflow traces within a specific session?▼

The best way to locate failed MLflow traces is to search for all traces matching a specific session ID and filter by failed status. This pinpoints execution issues for root cause analysis.

Does MLflow trace retrieval work with both CLI and Python API?▼

Yes, MLflow trace retrieval works with both CLI commands and a Python API. Both interfaces support querying trace data structures and spans for performance monitoring and error identification.

How do I query MLflow trace spans for performance monitoring?▼

Query MLflow trace spans by retrieving specific traces using their unique ID or searching by metadata and tags. This enables efficient querying of trace data structures to monitor execution performance.