A procedure for running StepCOVNet training and MERT feature extraction on a Windows computer through WSL, the Windows feature that runs a Linux environment. It also describes how to use the available GPU safely.

Starting training, running overfit tests, extracting MERT features with CUDA, checking GPU availability, and following the project's one-job-at-a-time rules.

It prevents GPU jobs from being run in the wrong environment or overlapping in ways that can cause failures or resource conflicts.

How cpuguy96/StepCOVNet configures Cursor

Reuse

Borrowing it

Nothing to install: this file belongs to cpuguy96/StepCOVNet. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/cpuguy96/StepCOVNet/master/.cursor/skills/whats-next/SKILL.md

Clone the repo

git clone --depth 1 https://github.com/cpuguy96/StepCOVNet

Made for: Cursor.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/cpuguy96/stepcovnet/whats-next/github.svg)](https://agentmods.dev/skills/cpuguy96/stepcovnet/whats-next)

Your own site

<a href="https://agentmods.dev/skills/cpuguy96/stepcovnet/whats-next"><img src="https://agentmods.dev/badge/skills/cpuguy96/stepcovnet/whats-next/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/cpuguy96/stepcovnet/whats-next"><img src="https://agentmods.dev/badge/skills/cpuguy96/stepcovnet/whats-next.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 66 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 2,128 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector warn 7 Sept 2026

SkillSpector: 2 findings, up to high

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

high Rogue Agent · line 48
Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Fix: Prevent the skill from modifying its own code, SKILL.md, or configuration files. Treat skill files as read-only at runtime.
high Rogue Agent · line 131
Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Fix: Prevent the skill from modifying its own code, SKILL.md, or configuration files. Treat skill files as read-only at runtime.

How audits are shown

Origin original No closer match found in the catalogue.

Token cost

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 5.1	$0.00066	$0.02128
Opus 5.5	$0.00026	$0.00851
Sonnet 5.5	$0.00013	$0.00426
Haiku 4.5	$0.00007	$0.00213

Measured 2mo ago against content hash 40f8a589f48a, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

whats-next 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 2mo ago.

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.

.cursor/skills/whats-next/SKILL.md · 148 lines

How it starts

The opening of the file, as written. The whole thing — 148 lines — stays where its author put it; the contents beside it link to each section on GitHub.

What's next

Use when the user asks what to do now, what's next, where are we, status, or orientation — including repeated asks across sessions.

Also load this skill when the user says do it, go ahead, start it, or run it after a whats-next answer (see § "Do it" below).

Workflow (same turn)

Run status (repo root, Windows venv):
```
venv\Scripts\python.exe scripts/project_status.py
```
Use --json only when merging into another tool.
Read EXPERIMENT_LOG.md § Current phase — that compact block is the live scoreboard (Next action, Blockers, numbers). This skill does not store them.
If project_status.py lists Missing prerequisites, fold them into Before you start — do not suggest launching a run until blockers are addressed or the user accepts the risk.
If GPU lock: held, say what job holds it; do not start another WSL GPU train/MERT job.
Skim the latest EXP-… row in the experiment index only when status or Current phase references an incomplete run.
Reply using the Answer template below. One primary action — not a menu unless the user asked for options.
Always fill If you say "do it" with numbered, copy-paste-ready steps (commands, config paths, log names). This block is the contract for a follow-up do it.

Evidence before recommending a run

Do not put a new train/probe/loss in Now (or If you say "do it") unless you have concrete evidence it addresses the binding failure — not only that Current phase / a NOTE named it as a backlog idea.

Allowed as Now	Not enough alone
Cheap diagnostics (ablation numbers, patch-acc curves, NLL vs uniform, error-mode analysis on an existing ckpt)	“Next action” text that lists an untested lever
A one-knob change with a mechanism already supported by logged numbers	Recipe spam (another λ / STE / aux) after similar recipes failed
A fix that already has unit/offline proof in-repo	Speculative architecture (“attn-mass might help”) with no supporting measurement

Read the full file on GitHub · 148 lines

Changes

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.

2mo ago First seen · 148 lines · 66 tokens per session scan A 40f8a589f48a

Subscribe to this mod's changes

whats-next is a skill published in the GitHub repository cpuguy96/StepCOVNet (22 stars, last pushed 27d ago), licensed Apache-2.0. It adds 66 tokens to every session and 2,128 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-08-30.

Planning Project management cnn deep-learning games machine-learning neural-network stepmania

whats-next next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
whats-nextcpuguy96/StepCOVNet	Skill	22	27d ago	Apache-2.0	A	66	Cursor
flow-nexus-neuralruvnet/ruflo	Skill	81k	today	MIT	A	21	Claude Code
developing-with-streamlitstreamlit/streamlit	Skill	46k	today	Apache-2.0	A	146	Claude Code
physicsnemo-shard-tensorNVIDIA/physicsnemo	Skill	3.4k	today	Apache-2.0	A	152	Claude Code, Codex
physicsnemo-discoverNVIDIA/physicsnemo	Skill	3.4k	today	Apache-2.0	A	124	Claude Code, Codex
lepton-clileptonai/leptonai	Skill	2.8k	2d ago	Apache-2.0	A	122	Claude Code
Home Security AI BenchmarkSharpAI/DeepCamera	Skill	3.1k	1mo ago	MIT	A	17	Claude Code, Codex

StepCOVNet: Skill for Cursor