mat-md-monitors

Monitor ASE molecular dynamics stability and halt simulations when monitor conditions trigger.

144|21|Updated Jan 8, 2026
One-click install
npx skills add https://github.com/learningmatter-mit/AtomisticSkills --skill mat-md-monitors
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mat-md-monitors
Source: https://github.com/learningmatter-mit/AtomisticSkills/tree/main/.agents/skills/mat-md-monitors
Command: npx skills add https://github.com/learningmatter-mit/AtomisticSkills --skill mat-md-monitors

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents wasted compute and unreliable results by detecting molecular dynamics instabilities early and terminating simulations safely before trajectories diverge or thermodynamic/transport properties become meaningless.

Core Features & Use Cases

  • Real-time stability monitoring: Uses ASE callbacks integrated into the run_md tool for zero-latency stopping when monitors trigger (e.g., explosion, equilibration, overshoot, volume instability, or diffusion convergence).
  • Targeted convergence checks: Stops based on temperature/potential energy stabilization and diffusion diffusivity relative-error thresholds for a chosen specie.
  • Temperature quenching support: Implements linear temperature ramps to move from temperature to temperature_end over a specified number of steps using thermostat-compatible updates.

Quick Start

Use the mat-md-monitors skill to run an ASE molecular dynamics simulation with monitor=True and a monitor_type such as "explosion" and "equilibration" so unstable or non-converged trajectories stop automatically.

Frequently Asked Questions about mat-md-monitors

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

FAQPage Schema
How do I stop unstable molecular dynamics simulations before trajectories diverge?

You can stop unstable molecular dynamics simulations by using callback-based monitoring that halts runs immediately with a "stopped" status and clear stop_reason when explosion, overshoot, or volume instability conditions are triggered.

Can I monitor ASE equilibration and diffusion convergence during MLIP-driven MD?

Yes, you can monitor ASE equilibration and diffusion convergence during MLIP-driven MD by checking temperature, potential energy stabilization, and diffusivity relative-error thresholds to stop simulations automatically once converged.

What is the best way to prevent wasted compute in NVT or NPT production runs?

The best way to prevent wasted compute in NVT or NPT production runs is to apply real-time stability monitoring with zero-latency stopping, which terminates simulations safely before thermodynamic properties become meaningless.

Does this molecular dynamics monitoring support temperature quenching with thermostat updates?

Yes, molecular dynamics monitoring supports temperature quenching by implementing linear temperature ramps from a starting temperature to an end temperature over specified steps using thermostat-compatible updates.

When do I need early termination for diffusion-focused molecular dynamics tasks?

Early termination for diffusion-focused molecular dynamics tasks is needed when you require reliable transport property outputs, allowing you to stop runs immediately upon reaching diffusivity convergence thresholds.

Why does my molecular dynamics trajectory explode during MLIP simulations?

Molecular dynamics trajectories explode during MLIP simulations due to instabilities that cause thermodynamic properties to diverge, which you can detect and halt early using explosion monitor conditions.