Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/learningmatter-mit/atomisticskills/ml-mlip-speednpx skills add learningmatter-mit/AtomisticSkills --skill ml-mlip-speedgit clone --depth 1 https://github.com/learningmatter-mit/AtomisticSkillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/learningmatter-mit/atomisticskills/ml-mlip-speed)<a href="https://agentmods.dev/skills/learningmatter-mit/atomisticskills/ml-mlip-speed"><img src="https://agentmods.dev/badge/skills/learningmatter-mit/atomisticskills/ml-mlip-speed.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00021 | $0.00905 |
| Opus 5 | $0.00010 | $0.00452 |
| Sonnet 5 | $0.00004 | $0.00181 |
| Haiku 4.5 | $0.00002 | $0.00090 |
Grade A, and why
ml-mlip-speed scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MLIP Performance Benchmarking
Goal
Evaluate and compare the inference speed (latency) and memory consumption of various foundation MLIP models to determine their suitability for different simulation scales and timescales.
Benchmark Script
The benchmark_mlips.py script measures performance by running short MD simulations on NaCl supercells of varying sizes.
Usage
Run the script within the appropriate conda environment for the models being tested. The script automatically skips models not supported by the current environment.
Multi-Environment Benchmarking
Because different MLIPs require isolated Conda environments (e.g., mace-agent, matgl-agent, fairchem-agent), the benchmark results are built incrementally.
- Run the script in each environment: The script gracefully skips models whose libraries are missing while preserving and updating the central
speed_benchmark.yamlfile. - Consolidate: Run the script in any environment (that has
matplotlib) with the--only_plotflag to generate the combined graphs from the accumulated total data.
# Example: Running in different environments sequentially
/path/to/mace-python benchmark_mlips.py --output_dir results/
/path/to/matgl-python benchmark_mlips.py --output_dir results/
/path/to/fairchem-python benchmark_mlips.py --output_dir results/
# Generate final combined plots
python benchmark_mlips.py --only_plot --output_dir results/
Key Arguments:
--models: List of model names/checkpoints to benchmark.--providers: Corresponding providers (mace,matgl,fairchem).--output_dir: Directory to save results and plots.--max_atoms_limit: Maximum system size to test (default: 5000).--only_plot: Re-generate plots from an existingspeed_benchmark.yamlfile without running simulations.
Metrics Explained
- Inference Time / Atom (ms): The normalized time taken for a single force/energy calculation per atom. Converged values (for larger systems) provide the best comparison.
- Memory Usage / Atom (MB): The peak VRAM footprint per atom. Useful for predicting OOM (Out Of Memory) limits for large supercells.
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 75 lines · 21 tokens per session scan A bd50e4ffc6d4
ml-mlip-speed is a skill published in the GitHub repository learningmatter-mit/AtomisticSkills (158 stars, last pushed yesterday), licensed MIT. It adds 21 tokens to every session and 905 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
hugging-face-trackio
Track and visualize ML training experiments with Trackio. Use when logging metrics during training (Python API) or retrieving/analyzing logged metrics (CLI). Supports real-time dashboard visualization, HF Space syncing, and JSON output for automation.
pysr
Use when fitting equations to data with PySR or SymbolicRegression.jl, when a user wants an interpretable formula, symbolic model, scaling law, or empirical relation discovered from numeric data, or when debugging a PySR search that is slow, stuck, or giving poor equations.
cellxgene-census-query
Query CZ CELLxGENE Census (61M+ cells). Filter by cell type/tissue/disease, retrieve expression data, and integrate with scanpy/PyTorch for population-scale single-cell analysis. Use this skill when: (1) Querying single-cell expression data by cell type, tissue, or disease, (2) Exploring available single-cell datasets…
HomeSafe-Bench
VLM indoor safety hazard detection benchmark inspired by HomeSafeBench (arXiv 2509.23690).
depth-estimation
Real-time depth map privacy transforms using Depth Anything v2 (CoreML + PyTorch).
start-experiment
Start the autoresearch optimization loop for a specific model + lane. Resolves the hierarchical program.md (root → model → lane), asks the user for hardware (local TPU VM or GKE cluster of a specified TPU type + topology), discovers available clusters from .env/, checks occupancy with USERPREFIX-aware attribution…