skills.rest
lix965996-art
lix965996-artCOMMUNITY
Explore all skills by this author →
GitHub

monitor-experiment

Monitor remote screen sessions and collect JSON results from machine-learning experiments.

1|Updated May 14, 2026
Search Tags:#JSON parsing#experiment monitoring#SSH screen#results aggregation#training status#Weights & Biases#metric comparison
One-click install
npx skills add https://github.com/lix965996-art/MMM --skill monitor-experiment-lix965996-art
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: monitor-experiment
Source: https://github.com/lix965996-art/MMM/tree/main/resources/app/skills/monitor-experiment
Command: npx skills add https://github.com/lix965996-art/MMM --skill monitor-experiment-lix965996-art

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It eliminates the guesswork of whether experiments are still running and what the latest outcomes are, by consolidating live screen progress and produced result artifacts into a clear summary.

Core Features & Use Cases

  • Progress monitoring: Checks which screen sessions are active to determine whether training is ongoing.
  • Result collection: Pulls recent output from screen sessions and inspects result directories for JSON outputs to surface the latest numeric findings.
  • Experiment interpretation & reporting: Compares current metrics against a baseline, flags anomalies (e.g., divergence or NaNs), and proposes next steps; optionally sends a Feishu notification when configured.

Quick Start

Ask the skill to monitor your experiment and summarize the latest results using your server alias or screen name as the argument.

Frequently Asked Questions about monitor-experiment

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

FAQPage Schema
How do I monitor running machine learning training status across remote SSH screen sessions?▼

To monitor training status, the skill checks active SSH screen sessions via hardcopy extraction and summarizes whether your machine learning experiments are still running or have completed.

How do I aggregate JSON result artifacts and compare metrics for multiple experiments?▼

You can collect JSON result artifacts and compare current metrics against a baseline by having the skill inspect your results directories to surface the latest numeric findings.

Can I parse Weights & Biases metrics during experiment monitoring?▼

Yes, Weights & Biases metrics can be optionally parsed and included in the experiment monitoring summary when enabled in your CLAUDE.md configuration file.

What is needed to check if my training runs are still active using screen session monitoring?▼

You need SSH access to the remote server to run screen listing and hardcopy extraction, which allows the monitor to determine if your training workflows are ongoing.

How do I detect training anomalies like divergence or NaNs in my experiment results?▼

The skill flags anomalies such as divergence or NaNs by comparing current metrics against a baseline and interpreting the collected JSON outputs from your experiment directories.

Does experiment monitoring support sending automated Feishu notifications for results?▼

Yes, the skill can optionally send a Feishu notification with your experiment summary and metric reporting when the notification feature is configured.

Related Skills

Explore More Data & Analytics Skills→
jkfeejkfee

monitor-experiment

Monitors running experiments and collects results via SSH, screen sessions, wandb workflows, and remote JSON retrieval.

Community
Intermediate
MACJONSWMACJONSW

monitor-experiment

Monitor remote experiments via SSH and collect results from logs.

Community
Intermediate
DifanaDAPDifanaDAP

weights-and-biases

Automate ML experiment tracking and hyperparameter sweeps via the Python API.

Community
Advanced
wanghao9610wanghao9610
39

star-expt-analyst

Audit AI experiment runs by verifying artifacts, logs, and metrics against plan criteria.

Community
Advanced