Home /Claude Skills /wandb-skill
Recommended

Wandb Skill

Say goodbye to manually organizing AI experiment data
AI experiments? Just ask Claude
Core Principle:
This skill turns your Claude into an instant AI experiment analyst! It automatically queries W&B experiments, analyzes Weave traces, and generates reports - like having a data scientist on call, without writing a single line of code.
KEY FEATURES
01Experiment Query
Get all experiment records with one click, 10x faster than manual lookup
02Trace Analysis
Auto-detect errors and latency bottlenecks
03Project Overview
See all team projects at a glance
04Smart Reporting
Auto-generate visualized weekly training reports
github.com/guangzhaoli/wandb-skill
data-ai·guangzhaoli·2026-01-29·1·🔱 0
Curated by agent-skills.cc
Installation
Download
HTTPS
git clone https://github.com/guangzhaoli/wandb-skill.git
SSH
git clone [email protected]:guangzhaoli/wandb-skill.git
GitHub CLI
gh repo clone guangzhaoli/wandb-skill
FAQ
Q: What are the installation steps for Wandb Skill Agent Skills?
1.Input Request: Describe your query in natural language
2.Auto Parse: Claude identifies intent and selects script
3.Execute Query: Calls W&B API in background
4.Present Results: Returns structured analysis report
Q: What are the highlights of Wandb Skill Agent Skills?
  • No-code AI experiment analysis
  • Natural language queries
  • Auto-generated beautiful reports
  • Team collaboration data sharing
Q: What are the use cases for Wandb Skill Agent Skills?
  • Quickly check model training progress before standup
  • Troubleshoot inference service issues
  • Present weekly experiment results to boss
  • Compare performance across model versions
Q: What are the limitations of Wandb Skill Agent Skills?
  • Requires pre-configured W&B API key
  • Only supports basic query functions